add api cli

.env.example

@@ -1,41 +1,47 @@
 # =============================================================================
-# MEMORA ENVIRONMENT CONFIGURATION
+# HINDSIGHT ENVIRONMENT CONFIGURATION
 # =============================================================================
 # Copy this file to .env and update with your values
 # Both services (API and Control Plane) read from this single file
 
 # =============================================================================
-# API SERVICE (MEMORA_API_*)
+# API SERVICE (HINDSIGHT_API_*)
 # =============================================================================
 
 # Database
-MEMORA_API_DATABASE_URL=postgresql://memora:memora_dev@localhost:5432/memora
+# Use "pg0" to start an embedded PostgreSQL instance via pg0
+# Or provide a full connection URL for external PostgreSQL
+#HINDSIGHT_API_DATABASE_URL=postgresql://hindsight:hindsight_dev@localhost:5432/hindsight
+HINDSIGHT_API_DATABASE_URL=pg0
+
+# pg0 data directory (only used when HINDSIGHT_API_DATABASE_URL=pg0)
+# HINDSIGHT_API_PG0_DATA_DIR=/path/to/pg_data
 
 # LLM Provider: "openai", "groq", or "ollama"
-MEMORA_API_LLM_PROVIDER=groq
+HINDSIGHT_API_LLM_PROVIDER=groq
 
 # LLM Model (provider-specific)
-MEMORA_API_LLM_MODEL=openai/gpt-oss-20b
+HINDSIGHT_API_LLM_MODEL=openai/gpt-oss-20b
 
 # API Key (not needed for ollama)
-MEMORA_API_LLM_API_KEY=your_api_key_here
+HINDSIGHT_API_LLM_API_KEY=your_api_key_here
 
 # Optional: Custom base URL (for ollama or custom endpoints)
-# MEMORA_API_LLM_BASE_URL=http://localhost:11434/v1
+# HINDSIGHT_API_LLM_BASE_URL=http://localhost:11434/v1
 
 # API Server Configuration (optional)
-# MEMORA_API_HOST=0.0.0.0
-# MEMORA_API_PORT=8080
+# HINDSIGHT_API_HOST=0.0.0.0
+# HINDSIGHT_API_PORT=8888
 
-MEMORA_API_MCP_ENABLED=true
+HINDSIGHT_API_MCP_ENABLED=true
 
 # =============================================================================
-# CONTROL PLANE SERVICE (MEMORA_CP_*)
+# CONTROL PLANE SERVICE (HINDSIGHT_CP_*)
 # =============================================================================
 
 # Dataplane API URL (where the control plane connects to)
-MEMORA_CP_DATAPLANE_API_URL=http://localhost:8080
+HINDSIGHT_CP_DATAPLANE_API_URL=http://localhost:8888
 
 # Control Plane Server Configuration (optional)
-# MEMORA_CP_PORT=3000
-# MEMORA_CP_HOSTNAME=0.0.0.0
+# HINDSIGHT_CP_PORT=3000
+# HINDSIGHT_CP_HOSTNAME=0.0.0.0

.github/workflows/deploy-docs.yml

@@ -2,9 +2,10 @@ name: Deploy Docs to GitHub Pages
 
 on:
   push:
-    branches: [main]
+    branches: [main, renaming-pre-launch]
     paths:
-      - 'memora-docs/**'
+      - 'hindsight-docs/**'
+      - '.github/workflows/deploy-docs.yml'
   workflow_dispatch:
 
 permissions:
@@ -21,20 +22,19 @@ jobs:
     runs-on: ubuntu-latest
     defaults:
       run:
-        working-directory: memora-docs
+        working-directory: hindsight-docs
     steps:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4
         with:
           node-version: 20
           cache: npm
-          cache-dependency-path: memora-docs/package-lock.json
+          cache-dependency-path: hindsight-docs/package-lock.json
       - run: npm ci
       - run: npm run build
       - uses: actions/upload-pages-artifact@v3
         with:
-          path: memora-docs/build
-
+          path: hindsight-docs/build
   deploy:
     environment:
       name: github-pages

.github/workflows/release.yml

@@ -22,15 +22,15 @@ jobs:
       with:
         python-version-file: ".python-version"
 
-    - name: Build memora package
-      working-directory: ./memora
+    - name: Build hindsight package
+      working-directory: ./hindsight
       run: uv build
 
     - name: Upload artifacts
       uses: actions/upload-artifact@v4
       with:
-        name: python-memora-dist
-        path: memora/dist/*
+        name: python-hindsight-dist
+        path: hindsight/dist/*
         retention-days: 30
 
   build-rust-cli:
@@ -40,16 +40,16 @@ jobs:
         include:
           - os: ubuntu-latest
             target: x86_64-unknown-linux-gnu
-            artifact_name: memora
-            asset_name: memora-linux-amd64
+            artifact_name: hindsight
+            asset_name: hindsight-linux-amd64
           - os: macos-latest
             target: x86_64-apple-darwin
-            artifact_name: memora
-            asset_name: memora-darwin-amd64
+            artifact_name: hindsight
+            asset_name: hindsight-darwin-amd64
           - os: macos-latest
             target: aarch64-apple-darwin
-            artifact_name: memora
-            asset_name: memora-darwin-arm64
+            artifact_name: hindsight
+            asset_name: hindsight-darwin-arm64
 
     steps:
     - uses: actions/checkout@v4
@@ -74,17 +74,17 @@ jobs:
     - name: Cache cargo build
       uses: actions/cache@v4
       with:
-        path: memora-cli/target
+        path: hindsight-cli/target
         key: ${{ runner.os }}-cargo-build-target-${{ hashFiles('**/Cargo.lock') }}
 
     - name: Build
-      working-directory: memora-cli
+      working-directory: hindsight-cli
       run: cargo build --release --target ${{ matrix.target }}
 
     - name: Prepare artifact
       run: |
         mkdir -p artifacts
-        cp memora-cli/target/${{ matrix.target }}/release/${{ matrix.artifact_name }} artifacts/${{ matrix.asset_name }}
+        cp hindsight-cli/target/${{ matrix.target }}/release/${{ matrix.artifact_name }} artifacts/${{ matrix.asset_name }}
         chmod +x artifacts/${{ matrix.asset_name }}
 
     - name: Upload artifacts
@@ -135,7 +135,7 @@ jobs:
       id: meta
       uses: docker/metadata-action@v5
       with:
-        images: ghcr.io/${{ github.repository_owner }}/memora-${{ matrix.component }}
+        images: ghcr.io/${{ github.repository_owner }}/hindsight-${{ matrix.component }}
         tags: |
           type=semver,pattern={{version}},value=${{ steps.get_version.outputs.VERSION }}
           type=semver,pattern={{major}}.{{minor}},value=${{ steps.get_version.outputs.VERSION }}
@@ -179,11 +179,11 @@ jobs:
 
     - name: Lint Helm chart
       run: |
-        helm lint helm/memora
+        helm lint helm/hindsight
 
     - name: Package Helm chart
       run: |
-        helm package helm/memora --destination ./helm-packages
+        helm package helm/hindsight --destination ./helm-packages
 
     - name: Upload Helm chart artifact
       uses: actions/upload-artifact@v4
@@ -208,26 +208,26 @@ jobs:
     - name: Download Python package
       uses: actions/download-artifact@v4
       with:
-        name: python-memora-dist
-        path: ./artifacts/python-memora-dist
+        name: python-hindsight-dist
+        path: ./artifacts/python-hindsight-dist
 
     - name: Download Rust CLI (Linux)
       uses: actions/download-artifact@v4
       with:
-        name: rust-cli-memora-linux-amd64
-        path: ./artifacts/rust-cli-memora-linux-amd64
+        name: rust-cli-hindsight-linux-amd64
+        path: ./artifacts/rust-cli-hindsight-linux-amd64
 
     - name: Download Rust CLI (macOS Intel)
       uses: actions/download-artifact@v4
       with:
-        name: rust-cli-memora-darwin-amd64
-        path: ./artifacts/rust-cli-memora-darwin-amd64
+        name: rust-cli-hindsight-darwin-amd64
+        path: ./artifacts/rust-cli-hindsight-darwin-amd64
 
     - name: Download Rust CLI (macOS ARM)
       uses: actions/download-artifact@v4
       with:
-        name: rust-cli-memora-darwin-arm64
-        path: ./artifacts/rust-cli-memora-darwin-arm64
+        name: rust-cli-hindsight-darwin-arm64
+        path: ./artifacts/rust-cli-hindsight-darwin-arm64
 
     - name: Download Helm chart
       uses: actions/download-artifact@v4
@@ -239,11 +239,11 @@ jobs:
       run: |
         mkdir -p release-assets
         # Python package
-        cp artifacts/python-memora-dist/* release-assets/
+        cp artifacts/python-hindsight-dist/* release-assets/
         # Rust CLI binaries
-        cp artifacts/rust-cli-memora-linux-amd64/memora-linux-amd64 release-assets/
-        cp artifacts/rust-cli-memora-darwin-amd64/memora-darwin-amd64 release-assets/
-        cp artifacts/rust-cli-memora-darwin-arm64/memora-darwin-arm64 release-assets/
+        cp artifacts/rust-cli-hindsight-linux-amd64/hindsight-linux-amd64 release-assets/
+        cp artifacts/rust-cli-hindsight-darwin-amd64/hindsight-darwin-amd64 release-assets/
+        cp artifacts/rust-cli-hindsight-darwin-arm64/hindsight-darwin-arm64 release-assets/
         # Helm chart
         cp artifacts/helm-chart/*.tgz release-assets/
 
@@ -251,68 +251,68 @@ jobs:
       id: release_notes
       run: |
         cat << EOF > release-notes.md
-        # Memora v${{ steps.get_version.outputs.VERSION }}
+        # Hindsight v${{ steps.get_version.outputs.VERSION }}
 
         ## 📦 Release Artifacts
 
         ### Python Package
-        - \`memora-${{ steps.get_version.outputs.VERSION }}-py3-none-any.whl\`
-        - \`memora-${{ steps.get_version.outputs.VERSION }}.tar.gz\`
+        - \`hindsight-${{ steps.get_version.outputs.VERSION }}-py3-none-any.whl\`
+        - \`hindsight-${{ steps.get_version.outputs.VERSION }}.tar.gz\`
 
         ### CLI Binaries
-        - \`memora-linux-amd64\` - Linux x86_64
-        - \`memora-darwin-amd64\` - macOS Intel
-        - \`memora-darwin-arm64\` - macOS Apple Silicon
+        - \`hindsight-linux-amd64\` - Linux x86_64
+        - \`hindsight-darwin-amd64\` - macOS Intel
+        - \`hindsight-darwin-arm64\` - macOS Apple Silicon
 
         ### Helm Chart
-        - \`memora-${{ steps.get_version.outputs.VERSION }}.tgz\`
+        - \`hindsight-${{ steps.get_version.outputs.VERSION }}.tgz\`
 
         ### Docker Images
         Docker images are published to GitHub Container Registry:
-        - \`ghcr.io/${{ github.repository_owner }}/memora-api:${{ steps.get_version.outputs.VERSION }}\`
-        - \`ghcr.io/${{ github.repository_owner }}/memora-control-plane:${{ steps.get_version.outputs.VERSION }}\`
+        - \`ghcr.io/${{ github.repository_owner }}/hindsight-api:${{ steps.get_version.outputs.VERSION }}\`
+        - \`ghcr.io/${{ github.repository_owner }}/hindsight-control-plane:${{ steps.get_version.outputs.VERSION }}\`
 
         ## 🚀 Installation
 
         ### Python Package
         \`\`\`bash
-        pip install memora==${{ steps.get_version.outputs.VERSION }}
+        pip install hindsight==${{ steps.get_version.outputs.VERSION }}
         \`\`\`
 
         ### CLI
         \`\`\`bash
         # macOS (Apple Silicon)
-        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/memora-darwin-arm64 -o memora
-        chmod +x memora
-        sudo mv memora /usr/local/bin/
+        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/hindsight-darwin-arm64 -o hindsight
+        chmod +x hindsight
+        sudo mv hindsight /usr/local/bin/
 
         # macOS (Intel)
-        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/memora-darwin-amd64 -o memora
-        chmod +x memora
-        sudo mv memora /usr/local/bin/
+        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/hindsight-darwin-amd64 -o hindsight
+        chmod +x hindsight
+        sudo mv hindsight /usr/local/bin/
 
         # Linux
-        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/memora-linux-amd64 -o memora
-        chmod +x memora
-        sudo mv memora /usr/local/bin/
+        curl -L https://github.com/${{ github.repository }}/releases/download/v${{ steps.get_version.outputs.VERSION }}/hindsight-linux-amd64 -o hindsight
+        chmod +x hindsight
+        sudo mv hindsight /usr/local/bin/
         \`\`\`
 
         ### Helm Chart
         \`\`\`bash
-        helm install memora memora-${{ steps.get_version.outputs.VERSION }}.tgz
+        helm install hindsight hindsight-${{ steps.get_version.outputs.VERSION }}.tgz
         \`\`\`
 
         ### Docker
         \`\`\`bash
         # Pull API image
-        docker pull ghcr.io/${{ github.repository_owner }}/memora-api:${{ steps.get_version.outputs.VERSION }}
+        docker pull ghcr.io/${{ github.repository_owner }}/hindsight-api:${{ steps.get_version.outputs.VERSION }}
 
         # Pull Control Plane image
-        docker pull ghcr.io/${{ github.repository_owner }}/memora-control-plane:${{ steps.get_version.outputs.VERSION }}
+        docker pull ghcr.io/${{ github.repository_owner }}/hindsight-control-plane:${{ steps.get_version.outputs.VERSION }}
 
         # Or use latest
-        docker pull ghcr.io/${{ github.repository_owner }}/memora-api:latest
-        docker pull ghcr.io/${{ github.repository_owner }}/memora-control-plane:latest
+        docker pull ghcr.io/${{ github.repository_owner }}/hindsight-api:latest
+        docker pull ghcr.io/${{ github.repository_owner }}/hindsight-control-plane:latest
         \`\`\`
         EOF
         cat release-notes.md
@@ -333,7 +333,7 @@ jobs:
         echo "# Release v${{ steps.get_version.outputs.VERSION }} Published Successfully" >> $GITHUB_STEP_SUMMARY
         echo "" >> $GITHUB_STEP_SUMMARY
         echo "## 📦 Components" >> $GITHUB_STEP_SUMMARY
-        echo "- ✅ Python package (memora)" >> $GITHUB_STEP_SUMMARY
+        echo "- ✅ Python package (hindsight)" >> $GITHUB_STEP_SUMMARY
         echo "- ✅ Rust CLI (Linux amd64, macOS amd64, macOS arm64)" >> $GITHUB_STEP_SUMMARY
         echo "- ✅ Docker images (API, Control Plane)" >> $GITHUB_STEP_SUMMARY
         echo "- ✅ Helm chart" >> $GITHUB_STEP_SUMMARY

.github/workflows/test.yml

@@ -16,7 +16,7 @@ jobs:
         env:
           POSTGRES_USER: postgres
           POSTGRES_PASSWORD: postgres
-          POSTGRES_DB: memora_test
+          POSTGRES_DB: hindsight_test
         options: >-
           --health-cmd pg_isready
           --health-interval 10s
@@ -26,10 +26,10 @@ jobs:
           - 5432:5432
 
     env:
-      MEMORA_API_DATABASE_URL: postgresql://postgres:postgres@localhost:5432/memora_test
-      MEMORA_API_LLM_PROVIDER: groq
-      MEMORA_API_LLM_API_KEY: ${{ secrets.GROQ_API_KEY }}
-      MEMORA_API_LLM_MODEL: openai/gpt-oss-20b
+      HINDSIGHT_API_DATABASE_URL: postgresql://postgres:postgres@localhost:5432/hindsight_test
+      HINDSIGHT_API_LLM_PROVIDER: groq
+      HINDSIGHT_API_LLM_API_KEY: ${{ secrets.GROQ_API_KEY }}
+      HINDSIGHT_API_LLM_MODEL: openai/gpt-oss-20b
 
     steps:
     - uses: actions/checkout@v4
@@ -48,9 +48,9 @@ jobs:
       run: uv sync --extra test
 
     - name: Run migrations
-      working-directory: ./memora
+      working-directory: ./hindsight
       run: |
         uv run alembic upgrade head
 
     - name: Run tests
-      run: uv run pytest memora/tests -v
+      run: uv run pytest hindsight/tests -v

CLAUDE.md

@@ -1,8 +0,0 @@
-# Documentation
-Do not write any markdown file, just write the code.
-
-# Workflow
-- After your changes, make sure everything is working fine by running the tests.
-- keep the readme.md architecture section up to date when you change the implementation
-- when changing an implemetation, do not keep the old one as fallback
-- to run test, use uv run pytest tests

README.md

@@ -1,64 +1,57 @@
-# Memora
+# Hindsight
 
 **Long-term memory for AI agents.**
 
-AI assistants forget everything between sessions. Memora fixes that with a memory system that handles temporal reasoning, entity connections, and personality-aware responses.
+AI assistants forget everything between sessions. Hindsight fixes that with a memory system that handles temporal reasoning, entity connections, and personality-aware responses.
 
-## Why Memora?
+## Why Hindsight?
 
 - **Temporal queries** — "What did Alice do last spring?" requires more than vector search
 - **Entity connections** — Knowing "Alice works at Google" + "Google is in Mountain View" = "Alice works in Mountain View"
 - **Agent opinions** — Agents form and recall beliefs with confidence scores
 - **Personality** — Big Five traits influence how agents process and respond to information
 
-## 5-Minute Setup
+## 60-seconds step
 
-### 1. Start the server
+
+### 1. Install the Hindsight All package (client + API)
 
 ```bash
-# Clone and start with Docker
-git clone https://github.com/anthropics/memora.git
-cd memora/docker
-./start.sh
+pip install hindsight-all
 ```
 
-Server runs at `http://localhost:8080`
-
-### 2. Install the Python client
-
+### 2. Import your OpenAI API key
 ```bash
-pip install memora-client
+export OPENAI_API_KEY=xx
 ```
 
-### 3. Use it
+### 3. Run embedded server and client
 
 ```python
-from memora_client import Memora
+import os
+from hindsight import HindsightServer, HindsightClient
 
-client = Memora(base_url="http://localhost:8080")
+with HindsightServer(llm_provider="openai", llm_model="gpt-5.1-mini", llm_api_key=os.environ["OPENAI_API_KEY"]) as server:
+    client = HindsightClient(base_url=server.url)
 
-# Store memories
-client.store(agent_id="my-agent", content="Alice works at Google")
-client.store(agent_id="my-agent", content="Bob prefers Python over JavaScript")
-
-# Search memories
-results = client.search(agent_id="my-agent", query="What does Alice do?")
-for r in results:
-    print(f"{r['text']} ({r['weight']:.2f})")
-
-# Generate personality-aware responses
-answer = client.think(agent_id="my-agent", query="Tell me about Alice")
-print(answer["text"])
+    client.put(agent_id="my-agent", content="Alice works at Google")
+    client.put(agent_id="my-agent", content="Bob prefers Python over JavaScript")
+    
+    client.search(agent_id="my-agent", query="What does Alice do?")
+    
+    client.think(agent_id="my-agent", query="Tell me about Alice")
 ```
 
+
+
 ## Documentation
 
-Full documentation: [memora-docs](./memora-docs)
+Full documentation: [hindsight-docs](./hindsight-docs)
 
-- [Architecture](./memora-docs/docs/developer/architecture.md) — How ingestion, storage, and retrieval work
-- [Python Client](./memora-docs/docs/sdks/python.md) — Full API reference
-- [API Reference](./memora-docs/docs/api-reference/index.md) — REST API endpoints
-- [Personality](./memora-docs/docs/developer/personality.md) — Big Five traits and opinion formation
+- [Architecture](./hindsight-docs/docs/developer/architecture.md) — How ingestion, storage, and retrieval work
+- [Python Client](./hindsight-docs/docs/sdks/python.md) — Full API reference
+- [API Reference](./hindsight-docs/docs/api-reference/index.md) — REST API endpoints
+- [Personality](./hindsight-docs/docs/developer/personality.md) — Big Five traits and opinion formation
 
 ## License
 

RELEASE.md

@@ -6,7 +6,7 @@
 
 ```bash
 uv sync
-cd memora-dev
+cd hindsight-dev
 uv run generate-openapi
 cd ..
 ```
@@ -24,7 +24,7 @@ This regenerates Python and TypeScript clients from `openapi.json`.
 ### 3. Commit Everything
 
 ```bash
-git add openapi.json memora-clients/
+git add openapi.json hindsight-clients/
 git commit -m "Update OpenAPI spec and regenerate clients"
 ```
 
@@ -47,7 +47,7 @@ This will:
 ### Publish Python Client to PyPI
 
 ```bash
-cd memora-clients/python
+cd hindsight-clients/python
 uv build
 uv publish
 ```
@@ -55,7 +55,7 @@ uv publish
 ### Publish TypeScript Client to NPM
 
 ```bash
-cd memora-clients/typescript
+cd hindsight-clients/typescript
 npm install
 npm run build
 npm publish --access public
@@ -65,7 +65,7 @@ npm publish --access public
 
 ## Pre-Release Checklist
 
-- [ ] Tests passing: `cd memora && uv run pytest tests`
+- [ ] Tests passing: `cd hindsight-api && uv run pytest tests`
 - [ ] No uncommitted changes: `git status`
 - [ ] On `main` branch
 
@@ -98,7 +98,7 @@ git status
 ```
 
 **GitHub Actions failed:**
-- Check: https://github.com/nicoloboschi/memora/actions
+- Check: https://github.com/vectorize-io/hindsight/actions
 - Re-run failed jobs or fix and release new patch version
 
 **Rollback:**
@@ -116,13 +116,13 @@ git push
 ```bash
 # Full release workflow
 uv sync
-cd memora-dev && uv run generate-openapi && cd ..
+cd hindsight-dev && uv run generate-openapi && cd ..
 ./scripts/generate-clients.sh
-git add openapi.json memora-clients/
+git add openapi.json hindsight-clients/
 git commit -m "Update OpenAPI spec and regenerate clients"
 ./scripts/release.sh 0.0.6
 
 # After GH Actions complete:
-cd memora-clients/python && uv build && uv publish
+cd hindsight-clients/python && uv build && uv publish
 cd ../typescript && npm run build && npm publish --access public
 ```

docker/README.md

@@ -1,175 +0,0 @@
-# Memora Docker Setup
-
-Complete Docker Compose setup for running all Memora services locally.
-
-## Services
-
-This setup includes:
-- **PostgreSQL** with pgvector extension (port 5432)
-- **API Service** - FastAPI backend (port 8080)
-- **Control Plane** - Next.js web UI (port 3000)
-
-## Quick Start
-
-1. **Configure environment variables:**
-   ```bash
-   cp .env.example .env
-   # Edit .env and set your API keys
-   ```
-
-2. **Start all services:**
-   ```bash
-   ./start.sh
-   ```
-
-3. **Access the services:**
-   - Control Plane: http://localhost:3000
-   - API: http://localhost:8080
-   - PostgreSQL: localhost:5432
-
-## Scripts
-
-### `./start.sh`
-Build and start all services. Waits for all services to be healthy.
-
-### `./stop.sh`
-Stop all services (keeps data).
-
-### `./clean.sh`
-Stop all services and remove all data (destructive).
-
-### `./logs.sh [service]`
-View logs for all services or a specific service:
-```bash
-./logs.sh              # All services
-./logs.sh api          # API only
-./logs.sh postgres     # PostgreSQL only
-./logs.sh control-plane # Control plane only
-```
-
-## Manual Docker Compose Commands
-
-```bash
-# Start services
-docker-compose up -d
-
-# Stop services
-docker-compose down
-
-# Rebuild and start
-docker-compose up --build -d
-
-# View logs
-docker-compose logs -f
-
-# Remove everything including data
-docker-compose down -v
-```
-
-## Database
-
-### Connection Info
-- **Host:** localhost
-- **Port:** 5432
-- **Database:** memora
-- **User:** memora
-- **Password:** memora_dev
-
-### Migrations
-
-Database migrations run automatically when the API service starts. The API uses Alembic to:
-1. Check the current schema version
-2. Run any pending migrations
-3. Initialize the database if it's empty
-
-Extensions (pgvector, uuid-ossp) are created automatically by the first migration.
-
-## Environment Variables
-
-Required in `.env` file:
-
-```bash
-# API Service Configuration
-MEMORA_API_DATABASE_URL=postgresql://memora:memora_dev@localhost:5432/memora
-MEMORA_API_LLM_PROVIDER=groq
-MEMORA_API_LLM_API_KEY=your-api-key-here
-MEMORA_API_LLM_MODEL=openai/gpt-oss-120b
-
-# Optional: Custom LLM endpoint
-# MEMORA_API_LLM_BASE_URL=http://localhost:11434/v1
-
-# Control Plane Configuration
-MEMORA_CP_DATAPLANE_API_URL=http://localhost:8080
-```
-
-## Troubleshooting
-
-### Services won't start
-Check logs for errors:
-```bash
-./logs.sh
-```
-
-### Database connection issues
-Ensure PostgreSQL is healthy:
-```bash
-docker exec memora-postgres pg_isready -U memora
-```
-
-### API won't connect to database
-Check if migrations ran successfully:
-```bash
-./logs.sh api
-```
-
-### Control plane can't reach API
-Verify the API is running:
-```bash
-curl http://localhost:8080/
-```
-
-### Reset everything
-```bash
-./clean.sh
-./start.sh
-```
-
-## Development
-
-### Rebuilding after code changes
-
-**API changes:**
-```bash
-docker-compose up --build -d api
-```
-
-**Control Plane changes:**
-```bash
-docker-compose up --build -d control-plane
-```
-
-### Accessing the database
-```bash
-docker exec -it memora-postgres psql -U memora -d memora
-```
-
-### Inspecting containers
-```bash
-docker-compose ps
-docker-compose exec api bash
-docker-compose exec control-plane sh
-```
-
-## Data Persistence
-
-PostgreSQL data is persisted in a Docker volume named `postgres_data`. This data survives container restarts but not `docker-compose down -v`.
-
-To backup data:
-```bash
-docker exec memora-postgres pg_dump -U memora memora > backup.sql
-```
-
-To restore data:
-```bash
-docker exec -i memora-postgres psql -U memora memora < backup.sql
-```

docker/api.Dockerfile

@@ -9,15 +9,15 @@ RUN apt-get update && apt-get install -y \
 WORKDIR /app
 
 # Copy only dependency files first for better caching
-COPY memora/pyproject.toml memora/README.md /app/memora/
-COPY memora/memora /app/memora/memora
-COPY memora/alembic /app/memora/alembic
+COPY hindsight-api/pyproject.toml hindsight-api/README.md /app/hindsight-api/
+COPY hindsight-api/hindsight_api /app/hindsight-api/hindsight_api
+COPY hindsight-api/alembic /app/hindsight-api/alembic
 
 # Install uv for faster dependency installation
 RUN pip install --no-cache-dir uv
 
 # Install Python dependencies to a virtual environment
-WORKDIR /app/memora
+WORKDIR /app/hindsight-api
 RUN uv venv /opt/venv && \
     . /opt/venv/bin/activate && \
     uv pip install --no-cache -e .
@@ -33,19 +33,19 @@ RUN apt-get update && apt-get install -y \
 
 # Copy virtual environment from builder
 COPY --from=builder /opt/venv /opt/venv
-COPY --from=builder /app/memora /app/memora
+COPY --from=builder /app/hindsight-api /app/hindsight-api
 
 # Set working directory
-WORKDIR /app/memora
+WORKDIR /app/hindsight-api
 
 # Expose API port
-EXPOSE 8080
+EXPOSE 8888
 
 # Set environment variables
 ENV PYTHONUNBUFFERED=1
-ENV DATABASE_URL=postgresql://memora:memora_dev@postgres:5432/memora
+ENV DATABASE_URL=postgresql://hindsight:hindsight_dev@postgres:5432/hindsight
 ENV PATH="/opt/venv/bin:$PATH"
-ENV PYTHONPATH=/app/memora
+ENV PYTHONPATH=/app/hindsight-api
 
 # Run the API server
-CMD ["python", "-m", "memora.web.server", "--host", "0.0.0.0", "--port", "8080"]
+CMD ["python", "-m", "hindsight_api.web.server", "--host", "0.0.0.0", "--port", "8888"]

docker/clean.sh

@@ -3,7 +3,7 @@ set -e
 
 cd "$(dirname "$0")"
 
-echo "🧹 Cleaning Memora Services"
+echo "🧹 Cleaning Services"
 echo "============================"
 echo ""
 echo "This will:"
@@ -20,7 +20,7 @@ fi
 
 echo ""
 echo "🗑️  Removing services and data..."
-docker-compose down -v
+docker compose down -v
 
 echo ""
 echo "✅ All services and data removed"

docker/control-plane.Dockerfile

@@ -31,9 +31,9 @@ COPY --from=builder --chown=nextjs:nodejs /app/.next/static ./.next/static
 
 USER nextjs
 
-EXPOSE 3000
+EXPOSE 9999
 
-ENV PORT=3000
+ENV PORT=9999
 ENV HOSTNAME="0.0.0.0"
 
 CMD ["node", "server.js"]

docker/docker-compose.yml

@@ -1,78 +1,78 @@
 services:
   postgres:
     image: pgvector/pgvector:pg16
-    container_name: memora-postgres
+    container_name: hindsight-postgres
     environment:
-      POSTGRES_USER: memora
-      POSTGRES_PASSWORD: memora_dev
-      POSTGRES_DB: memora
+      POSTGRES_USER: hindsight
+      POSTGRES_PASSWORD: hindsight_dev
+      POSTGRES_DB: hindsight
     ports:
       - "5432:5432"
     volumes:
       - postgres_data:/var/lib/postgresql/data
     healthcheck:
-      test: ["CMD-SHELL", "pg_isready -U memora"]
+      test: ["CMD-SHELL", "pg_isready -U hindsight"]
       interval: 5s
       timeout: 5s
       retries: 5
     networks:
-      - memora-network
+      - hindsight-network
 
   api:
     build:
       context: ..
       dockerfile: docker/api.Dockerfile
-    container_name: memora-api
+    container_name: hindsight-api
     environment:
-      MEMORA_API_DATABASE_URL: postgresql://memora:memora_dev@postgres:5432/memora
-      MEMORA_API_LLM_PROVIDER: ${MEMORA_API_LLM_PROVIDER:-groq}
-      MEMORA_API_LLM_API_KEY: ${MEMORA_API_LLM_API_KEY}
-      MEMORA_API_LLM_MODEL: ${MEMORA_API_LLM_MODEL:-openai/gpt-oss-120b}
-      MEMORA_API_LLM_BASE_URL: ${MEMORA_API_LLM_BASE_URL}
-      MEMORA_API_HOST: ${MEMORA_API_HOST:-0.0.0.0}
-      MEMORA_API_PORT: ${MEMORA_API_PORT:-8080}
+      HINDSIGHT_API_DATABASE_URL: postgresql://hindsight:hindsight_dev@postgres:5432/hindsight
+      HINDSIGHT_API_LLM_PROVIDER: ${HINDSIGHT_API_LLM_PROVIDER:-groq}
+      HINDSIGHT_API_LLM_API_KEY: ${HINDSIGHT_API_LLM_API_KEY}
+      HINDSIGHT_API_LLM_MODEL: ${HINDSIGHT_API_LLM_MODEL:-openai/gpt-oss-20b}
+      HINDSIGHT_API_LLM_BASE_URL: ${HINDSIGHT_API_LLM_BASE_URL}
+      HINDSIGHT_API_HOST: 0.0.0.0
+      HINDSIGHT_API_PORT: 8888
     ports:
-      - "8080:8080"
+      - "8888:8888"
     depends_on:
       postgres:
         condition: service_healthy
     healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:8080/api/v1/agents"]
+      test: ["CMD", "curl", "-f", "http://localhost:8888/api/v1/agents"]
       interval: 10s
       timeout: 5s
       retries: 5
       start_period: 30s
     networks:
-      - memora-network
+      - hindsight-network
     restart: unless-stopped
 
   control-plane:
     build:
-      context: ../memora-control-plane
+      context: ../hindsight-control-plane
       dockerfile: ../docker/control-plane.Dockerfile
-    container_name: memora-control-plane
+    container_name: hindsight-control-plane
     environment:
       NODE_ENV: production
-      MEMORA_CP_HOSTNAME: ${MEMORA_CP_HOSTNAME:-0.0.0.0}
-      MEMORA_CP_PORT: ${MEMORA_CP_PORT:-3000}
-      MEMORA_CP_DATAPLANE_API_URL: ${MEMORA_CP_DATAPLANE_API_URL:-http://api:8080}
+      HOSTNAME: 0.0.0.0
+      PORT: 9999
+      HINDSIGHT_CP_DATAPLANE_API_URL: http://api:8888
     ports:
-      - "3000:3000"
+      - "9999:9999"
     depends_on:
       api:
         condition: service_healthy
     healthcheck:
-      test: ["CMD", "wget", "--no-verbose", "--tries=1", "--spider", "http://localhost:3000/"]
+      test: ["CMD", "wget", "--no-verbose", "--tries=1", "--spider", "http://localhost:9999/"]
       interval: 10s
       timeout: 5s
       retries: 5
       start_period: 30s
     networks:
-      - memora-network
+      - hindsight-network
     restart: unless-stopped
 
 networks:
-  memora-network:
+  hindsight-network:
     driver: bridge
 
 volumes:

docker/logs.sh

@@ -8,9 +8,9 @@ SERVICE=$1
 if [ -z "$SERVICE" ]; then
     echo "📋 Showing logs for all services..."
     echo ""
-    docker-compose logs -f
+    docker compose logs -f
 else
     echo "📋 Showing logs for $SERVICE..."
     echo ""
-    docker-compose logs -f "$SERVICE"
+    docker compose logs -f "$SERVICE"
 fi

docker/start.sh

@@ -3,7 +3,7 @@ set -e
 
 cd "$(dirname "$0")"
 
-echo "🚀 Starting Memora Services"
+echo "🚀 Starting Hindsight Services"
 echo "============================"
 echo ""
 
@@ -15,14 +15,14 @@ if [ ! -f ../.env ]; then
     cp ../.env.example ../.env
     echo ""
     echo "⚠️  Please edit .env and set your API keys:"
-    echo "   - MEMORA_API_LLM_API_KEY"
+    echo "   - HINDSIGHT_API_LLM_API_KEY"
     echo ""
     echo "Then run this script again."
     exit 1
 fi
 
 echo "📦 Building and starting services..."
-docker-compose --env-file ../.env up --build -d
+docker compose --env-file ../.env up --build -d
 
 echo ""
 echo "⏳ Waiting for services to be healthy..."
@@ -30,21 +30,21 @@ echo ""
 
 # Wait for PostgreSQL
 echo "  Waiting for PostgreSQL..."
-until docker exec memora-postgres pg_isready -U memora > /dev/null 2>&1; do
+until docker exec hindsight-postgres pg_isready -U hindsight > /dev/null 2>&1; do
   sleep 1
 done
 echo "  ✅ PostgreSQL is ready"
 
 # Wait for API
 echo "  Waiting for API..."
-until curl -f http://localhost:8080/api/v1/agents > /dev/null 2>&1; do
+until curl -f http://localhost:8888/api/v1/agents > /dev/null 2>&1; do
   sleep 2
 done
 echo "  ✅ API is ready"
 
 # Wait for Control Plane
 echo "  Waiting for Control Plane..."
-until curl -f http://localhost:3000 > /dev/null 2>&1; do
+until curl -f http://localhost:9999 > /dev/null 2>&1; do
   sleep 2
 done
 echo "  ✅ Control Plane is ready"
@@ -53,12 +53,12 @@ echo ""
 echo "✅ All services are running!"
 echo ""
 echo "📊 Service URLs:"
-echo "   Control Plane: http://localhost:3000"
-echo "   API:           http://localhost:8080"
+echo "   Control Plane: http://localhost:9999"
+echo "   API:           http://localhost:8888"
 echo "   PostgreSQL:    localhost:5432"
 echo ""
 echo "🔍 View logs:"
-echo "   docker-compose logs -f"
+echo "   docker compose logs -f"
 echo ""
 echo "🛑 Stop services:"
 echo "   ./stop.sh"

docker/stop.sh

@@ -3,15 +3,15 @@ set -e
 
 cd "$(dirname "$0")"
 
-echo "🛑 Stopping Memora Services"
+echo "🛑 Stopping Services"
 echo "============================"
 echo ""
 
-docker-compose down
+docker compose down
 
 echo ""
 echo "✅ All services stopped"
 echo ""
 echo "💡 To remove data volumes as well, run:"
-echo "   docker-compose down -v"
+echo "   docker compose down -v"
 echo ""

helm/INSTALL.txt

@@ -1,4 +1,4 @@
-MEMORA HELM CHART INSTALLATION GUIDE
+HINDSIGHT HELM CHART INSTALLATION GUIDE
 =====================================
 
 PREREQUISITES
@@ -13,42 +13,42 @@ BASIC INSTALLATION
 
 1. Install with default values (requires external PostgreSQL):
 
-   helm install memora ./memora \
+   helm install hindsight ./hindsight \
      --set postgresql.external.host=your-postgres-host \
      --set postgresql.external.password=your-password \
      --set api.secrets.MEMORY_LLM_API_KEY=your-api-key
 
 2. Install with custom values file:
 
-   helm install memora ./memora -f memora/values-production.yaml
+   helm install hindsight ./hindsight -f hindsight/values-production.yaml
 
 3. Install in a specific namespace:
 
-   kubectl create namespace memora
-   helm install memora ./memora -n memora
+   kubectl create namespace hindsight
+   helm install hindsight ./hindsight -n hindsight
 
 CONFIGURATION OPTIONS
 ---------------------
 
 Development setup (using values-development.yaml):
-   helm install memora ./memora -f memora/values-development.yaml
+   helm install hindsight ./hindsight -f hindsight/values-development.yaml
 
 Production setup (using values-production.yaml):
-   helm install memora ./memora -f memora/values-production.yaml
+   helm install hindsight ./hindsight -f hindsight/values-production.yaml
 
 Custom LLM provider:
-   helm install memora ./memora \
+   helm install hindsight ./hindsight \
      --set api.env.MEMORY_LLM_PROVIDER=openai \
      --set api.env.MEMORY_LLM_MODEL=gpt-4 \
      --set api.secrets.MEMORY_LLM_API_KEY=sk-your-key
 
 Enable ingress:
-   helm install memora ./memora \
+   helm install hindsight ./hindsight \
      --set ingress.enabled=true \
-     --set ingress.hosts[0].host=memora.example.com
+     --set ingress.hosts[0].host=hindsight.example.com
 
 Enable autoscaling:
-   helm install memora ./memora \
+   helm install hindsight ./hindsight \
      --set autoscaling.enabled=true \
      --set autoscaling.minReplicas=2 \
      --set autoscaling.maxReplicas=10
@@ -57,43 +57,43 @@ UPGRADE
 -------
 
 Upgrade existing installation:
-   helm upgrade memora ./memora
+   helm upgrade hindsight ./hindsight
 
 Upgrade with new values:
-   helm upgrade memora ./memora -f memora/values-production.yaml
+   helm upgrade hindsight ./hindsight -f hindsight/values-production.yaml
 
 UNINSTALL
 ---------
 
 Remove the Helm release:
-   helm uninstall memora
+   helm uninstall hindsight
 
 Remove with namespace:
-   helm uninstall memora -n memora
+   helm uninstall hindsight -n hindsight
 
 TESTING
 -------
 
 Test the installation with dry-run:
-   helm install memora ./memora --dry-run --debug
+   helm install hindsight ./hindsight --dry-run --debug
 
 Validate templates:
-   helm template memora ./memora
+   helm template hindsight ./hindsight
 
 Lint the chart:
-   helm lint ./memora
+   helm lint ./hindsight
 
 ACCESSING THE SERVICES
 ----------------------
 
 Port-forward control plane:
-   kubectl port-forward svc/memora-control-plane 3000:3000
+   kubectl port-forward svc/hindsight-control-plane 3000:3000
 
 Port-forward API:
-   kubectl port-forward svc/memora-api 8080:8080
+   kubectl port-forward svc/hindsight-api 8888:8888
 
 Get service URLs:
-   helm status memora
+   helm status hindsight
 
 DATABASE INITIALIZATION
 -----------------------
@@ -102,16 +102,16 @@ NOTE: Database migrations now run automatically when the API service starts.
       You typically don't need to run migrations manually.
 
 If you want to pre-initialize the database before deploying (optional):
-   kubectl run memora-init --rm -it --restart=Never \
-     --image=memora/api:latest \
-     --env="DATABASE_URL=postgresql://user:pass@host:5432/memora" \
-     -- python -c "from memora.migrations import run_migrations; run_migrations()"
+   kubectl run hindsight-init --rm -it --restart=Never \
+     --image=hindsight/api:latest \
+     --env="DATABASE_URL=postgresql://user:pass@host:5432/hindsight" \
+     -- python -c "from hindsight.migrations import run_migrations; run_migrations()"
 
 TROUBLESHOOTING
 ---------------
 
 Check pod status:
-   kubectl get pods -l app.kubernetes.io/name=memora
+   kubectl get pods -l app.kubernetes.io/name=hindsight
 
 View logs for API:
    kubectl logs -l app.kubernetes.io/component=api
@@ -123,8 +123,8 @@ Describe a pod:
    kubectl describe pod <pod-name>
 
 Check configuration:
-   kubectl get configmap memora-config -o yaml
-   kubectl get secret memora-secret -o yaml
+   kubectl get configmap hindsight-config -o yaml
+   kubectl get secret hindsight-secret -o yaml
 
 NOTES
 -----

helm/hindsight/Chart.yaml

@@ -0,0 +1,13 @@
+apiVersion: v2
+name: hindsight
+description: A Helm chart for Hindsight - temporal-semantic-entity memory system for AI agents
+type: application
+version: 0.0.7
+appVersion: "0.0.7"
+keywords:
+  - ai
+  - memory
+  - llm
+  - agents
+maintainers:
+  - name: Hindsight Team

helm/hindsight/templates/NOTES.txt

@@ -18,13 +18,13 @@ The application is accessible via the following URL(s):
 
 1. Get the Control Plane URL by running these commands:
 {{- if contains "NodePort" .Values.controlPlane.service.type }}
-  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "memora.fullname" . }}-control-plane)
+  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "hindsight.fullname" . }}-control-plane)
   export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
   echo "Control Plane URL: http://$NODE_IP:$NODE_PORT"
 {{- else if contains "LoadBalancer" .Values.controlPlane.service.type }}
   NOTE: It may take a few minutes for the LoadBalancer IP to be available.
-  You can watch the status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "memora.fullname" . }}-control-plane'
-  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "memora.fullname" . }}-control-plane --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
+  You can watch the status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "hindsight.fullname" . }}-control-plane'
+  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "hindsight.fullname" . }}-control-plane --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
   echo "Control Plane URL: http://$SERVICE_IP:{{ .Values.controlPlane.service.port }}"
 {{- else if contains "ClusterIP" .Values.controlPlane.service.type }}
   export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/component=control-plane,app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
@@ -35,19 +35,19 @@ The application is accessible via the following URL(s):
 
 2. Get the API URL by running these commands:
 {{- if contains "NodePort" .Values.api.service.type }}
-  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "memora.fullname" . }}-api)
+  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "hindsight.fullname" . }}-api)
   export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
   echo "API URL: http://$NODE_IP:$NODE_PORT"
 {{- else if contains "LoadBalancer" .Values.api.service.type }}
   NOTE: It may take a few minutes for the LoadBalancer IP to be available.
-  You can watch the status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "memora.fullname" . }}-api'
-  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "memora.fullname" . }}-api --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
+  You can watch the status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "hindsight.fullname" . }}-api'
+  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "hindsight.fullname" . }}-api --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
   echo "API URL: http://$SERVICE_IP:{{ .Values.api.service.port }}"
 {{- else if contains "ClusterIP" .Values.api.service.type }}
   export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/component=api,app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
   export CONTAINER_PORT=$(kubectl get pod --namespace {{ .Release.Namespace }} $POD_NAME -o jsonpath="{.spec.containers[0].ports[0].containerPort}")
-  echo "API URL: http://127.0.0.1:8080"
-  kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8080:$CONTAINER_PORT
+  echo "API URL: http://127.0.0.1:8888"
+  kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8888:$CONTAINER_PORT
 {{- end }}
 
 {{- end }}
@@ -62,10 +62,10 @@ Please ensure that:
 Database migrations run automatically when the API service starts.
 
 If you want to pre-initialize the database before deploying (optional):
-  kubectl run --namespace {{ .Release.Namespace }} memora-init --rm -it --restart=Never \
+  kubectl run --namespace {{ .Release.Namespace }} hindsight-init --rm -it --restart=Never \
     --image={{ .Values.api.image.repository }}:{{ .Values.api.image.tag }} \
-    --env="DATABASE_URL={{ include "memora.databaseUrl" . }}" \
-    -- python -c "from memora.migrations import run_migrations; run_migrations()"
+    --env="DATABASE_URL={{ include "hindsight.databaseUrl" . }}" \
+    -- python -c "from hindsight.migrations import run_migrations; run_migrations()"
 {{- end }}
 
-For more information, visit: https://github.com/yourusername/memora
+For more information, visit: https://github.com/yourusername/hindsight

helm/hindsight/templates/api-deployment.yaml

@@ -2,16 +2,16 @@
 apiVersion: apps/v1
 kind: Deployment
 metadata:
-  name: {{ include "memora.fullname" . }}-api
+  name: {{ include "hindsight.fullname" . }}-api
   labels:
-    {{- include "memora.api.labels" . | nindent 4 }}
+    {{- include "hindsight.api.labels" . | nindent 4 }}
 spec:
   {{- if not .Values.autoscaling.enabled }}
   replicas: {{ .Values.api.replicaCount }}
   {{- end }}
   selector:
     matchLabels:
-      {{- include "memora.api.selectorLabels" . | nindent 6 }}
+      {{- include "hindsight.api.selectorLabels" . | nindent 6 }}
   template:
     metadata:
       annotations:
@@ -21,10 +21,10 @@ spec:
         {{- toYaml . | nindent 8 }}
         {{- end }}
       labels:
-        {{- include "memora.api.selectorLabels" . | nindent 8 }}
+        {{- include "hindsight.api.selectorLabels" . | nindent 8 }}
     spec:
       {{- if .Values.serviceAccount.create }}
-      serviceAccountName: {{ include "memora.serviceAccountName" . }}
+      serviceAccountName: {{ include "hindsight.serviceAccountName" . }}
       {{- end }}
       securityContext:
         {{- toYaml .Values.podSecurityContext | nindent 8 }}
@@ -39,37 +39,37 @@ spec:
           containerPort: {{ .Values.api.service.targetPort }}
           protocol: TCP
         env:
-        - name: MEMORA_API_DATABASE_URL
-          value: {{ include "memora.databaseUrl" . | quote }}
+        - name: HINDSIGHT_API_DATABASE_URL
+          value: {{ include "hindsight.databaseUrl" . | quote }}
         {{- if not .Values.postgresql.enabled }}
         - name: POSTGRES_PASSWORD
           valueFrom:
             secretKeyRef:
-              name: {{ include "memora.fullname" . }}-secret
+              name: {{ include "hindsight.fullname" . }}-secret
               key: postgres-password
         {{- end }}
-        - name: MEMORA_API_LLM_PROVIDER
+        - name: HINDSIGHT_API_LLM_PROVIDER
           valueFrom:
             configMapKeyRef:
-              name: {{ include "memora.fullname" . }}-config
+              name: {{ include "hindsight.fullname" . }}-config
               key: llm-provider
-        - name: MEMORA_API_LLM_MODEL
+        - name: HINDSIGHT_API_LLM_MODEL
           valueFrom:
             configMapKeyRef:
-              name: {{ include "memora.fullname" . }}-config
+              name: {{ include "hindsight.fullname" . }}-config
               key: llm-model
-        {{- if and .Values.api.secrets (hasKey .Values.api.secrets "MEMORA_API_LLM_API_KEY") }}
-        - name: MEMORA_API_LLM_API_KEY
+        {{- if and .Values.api.secrets (hasKey .Values.api.secrets "HINDSIGHT_API_LLM_API_KEY") }}
+        - name: HINDSIGHT_API_LLM_API_KEY
           valueFrom:
             secretKeyRef:
-              name: {{ include "memora.fullname" . }}-secret
+              name: {{ include "hindsight.fullname" . }}-secret
               key: llm-api-key
         {{- end }}
-        {{- if and .Values.api.secrets (hasKey .Values.api.secrets "MEMORA_API_LLM_BASE_URL") }}
-        - name: MEMORA_API_LLM_BASE_URL
+        {{- if and .Values.api.secrets (hasKey .Values.api.secrets "HINDSIGHT_API_LLM_BASE_URL") }}
+        - name: HINDSIGHT_API_LLM_BASE_URL
           valueFrom:
             secretKeyRef:
-              name: {{ include "memora.fullname" . }}-secret
+              name: {{ include "hindsight.fullname" . }}-secret
               key: llm-base-url
         {{- end }}
         livenessProbe:

helm/hindsight/templates/api-service.yaml

@@ -2,9 +2,9 @@
 apiVersion: v1
 kind: Service
 metadata:
-  name: {{ include "memora.fullname" . }}-api
+  name: {{ include "hindsight.fullname" . }}-api
   labels:
-    {{- include "memora.api.labels" . | nindent 4 }}
+    {{- include "hindsight.api.labels" . | nindent 4 }}
 spec:
   type: {{ .Values.api.service.type }}
   ports:
@@ -13,5 +13,5 @@ spec:
     protocol: TCP
     name: http
   selector:
-    {{- include "memora.api.selectorLabels" . | nindent 4 }}
+    {{- include "hindsight.api.selectorLabels" . | nindent 4 }}
 {{- end }}

helm/hindsight/templates/configmap.yaml

@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: {{ include "hindsight.fullname" . }}-config
+  labels:
+    {{- include "hindsight.labels" . | nindent 4 }}
+data:
+  # API configuration
+  llm-provider: {{ .Values.api.env.HINDSIGHT_API_LLM_PROVIDER | quote }}
+  llm-model: {{ .Values.api.env.HINDSIGHT_API_LLM_MODEL | quote }}
+
+  # Control plane configuration
+  node-env: {{ .Values.controlPlane.env.NODE_ENV | quote }}
+  hostname: {{ .Values.controlPlane.env.HINDSIGHT_CP_HOSTNAME | quote }}
+  control-plane-port: {{ .Values.controlPlane.env.HINDSIGHT_CP_PORT | quote }}

helm/hindsight/templates/controlplane-deployment.yaml

@@ -2,16 +2,16 @@
 apiVersion: apps/v1
 kind: Deployment
 metadata:
-  name: {{ include "memora.fullname" . }}-control-plane
+  name: {{ include "hindsight.fullname" . }}-control-plane
   labels:
-    {{- include "memora.controlPlane.labels" . | nindent 4 }}
+    {{- include "hindsight.controlPlane.labels" . | nindent 4 }}
 spec:
   {{- if not .Values.autoscaling.enabled }}
   replicas: {{ .Values.controlPlane.replicaCount }}
   {{- end }}
   selector:
     matchLabels:
-      {{- include "memora.controlPlane.selectorLabels" . | nindent 6 }}
+      {{- include "hindsight.controlPlane.selectorLabels" . | nindent 6 }}
   template:
     metadata:
       annotations:
@@ -20,10 +20,10 @@ spec:
         {{- toYaml . | nindent 8 }}
         {{- end }}
       labels:
-        {{- include "memora.controlPlane.selectorLabels" . | nindent 8 }}
+        {{- include "hindsight.controlPlane.selectorLabels" . | nindent 8 }}
     spec:
       {{- if .Values.serviceAccount.create }}
-      serviceAccountName: {{ include "memora.serviceAccountName" . }}
+      serviceAccountName: {{ include "hindsight.serviceAccountName" . }}
       {{- end }}
       securityContext:
         {{- toYaml .Values.podSecurityContext | nindent 8 }}
@@ -41,20 +41,20 @@ spec:
         - name: NODE_ENV
           valueFrom:
             configMapKeyRef:
-              name: {{ include "memora.fullname" . }}-config
+              name: {{ include "hindsight.fullname" . }}-config
               key: node-env
-        - name: MEMORA_CP_HOSTNAME
+        - name: HINDSIGHT_CP_HOSTNAME
           valueFrom:
             configMapKeyRef:
-              name: {{ include "memora.fullname" . }}-config
+              name: {{ include "hindsight.fullname" . }}-config
               key: hostname
-        - name: MEMORA_CP_PORT
+        - name: HINDSIGHT_CP_PORT
           valueFrom:
             configMapKeyRef:
-              name: {{ include "memora.fullname" . }}-config
+              name: {{ include "hindsight.fullname" . }}-config
               key: control-plane-port
-        - name: MEMORA_CP_DATAPLANE_API_URL
-          value: {{ include "memora.apiUrl" . | quote }}
+        - name: HINDSIGHT_CP_DATAPLANE_API_URL
+          value: {{ include "hindsight.apiUrl" . | quote }}
         livenessProbe:
           {{- toYaml .Values.controlPlane.livenessProbe | nindent 10 }}
         readinessProbe:

helm/hindsight/templates/controlplane-service.yaml

@@ -2,9 +2,9 @@
 apiVersion: v1
 kind: Service
 metadata:
-  name: {{ include "memora.fullname" . }}-control-plane
+  name: {{ include "hindsight.fullname" . }}-control-plane
   labels:
-    {{- include "memora.controlPlane.labels" . | nindent 4 }}
+    {{- include "hindsight.controlPlane.labels" . | nindent 4 }}
 spec:
   type: {{ .Values.controlPlane.service.type }}
   ports:
@@ -13,5 +13,5 @@ spec:
     protocol: TCP
     name: http
   selector:
-    {{- include "memora.controlPlane.selectorLabels" . | nindent 4 }}
+    {{- include "hindsight.controlPlane.selectorLabels" . | nindent 4 }}
 {{- end }}

helm/hindsight/templates/hpa.yaml

@@ -3,14 +3,14 @@
 apiVersion: autoscaling/v2
 kind: HorizontalPodAutoscaler
 metadata:
-  name: {{ include "memora.fullname" . }}-api
+  name: {{ include "hindsight.fullname" . }}-api
   labels:
-    {{- include "memora.api.labels" . | nindent 4 }}
+    {{- include "hindsight.api.labels" . | nindent 4 }}
 spec:
   scaleTargetRef:
     apiVersion: apps/v1
     kind: Deployment
-    name: {{ include "memora.fullname" . }}-api
+    name: {{ include "hindsight.fullname" . }}-api
   minReplicas: {{ .Values.autoscaling.minReplicas }}
   maxReplicas: {{ .Values.autoscaling.maxReplicas }}
   metrics:
@@ -34,14 +34,14 @@ spec:
 apiVersion: autoscaling/v2
 kind: HorizontalPodAutoscaler
 metadata:
-  name: {{ include "memora.fullname" . }}-control-plane
+  name: {{ include "hindsight.fullname" . }}-control-plane
   labels:
-    {{- include "memora.controlPlane.labels" . | nindent 4 }}
+    {{- include "hindsight.controlPlane.labels" . | nindent 4 }}
 spec:
   scaleTargetRef:
     apiVersion: apps/v1
     kind: Deployment
-    name: {{ include "memora.fullname" . }}-control-plane
+    name: {{ include "hindsight.fullname" . }}-control-plane
   minReplicas: {{ .Values.autoscaling.minReplicas }}
   maxReplicas: {{ .Values.autoscaling.maxReplicas }}
   metrics:

helm/hindsight/templates/ingress.yaml

@@ -2,9 +2,9 @@
 apiVersion: networking.k8s.io/v1
 kind: Ingress
 metadata:
-  name: {{ include "memora.fullname" . }}
+  name: {{ include "hindsight.fullname" . }}
   labels:
-    {{- include "memora.labels" . | nindent 4 }}
+    {{- include "hindsight.labels" . | nindent 4 }}
   {{- with .Values.ingress.annotations }}
   annotations:
     {{- toYaml . | nindent 4 }}
@@ -34,11 +34,11 @@ spec:
             backend:
               service:
                 {{- if eq .service "api" }}
-                name: {{ include "memora.fullname" $ }}-api
+                name: {{ include "hindsight.fullname" $ }}-api
                 port:
                   number: {{ $.Values.api.service.port }}
                 {{- else if eq .service "controlPlane" }}
-                name: {{ include "memora.fullname" $ }}-control-plane
+                name: {{ include "hindsight.fullname" $ }}-control-plane
                 port:
                   number: {{ $.Values.controlPlane.service.port }}
                 {{- end }}

helm/hindsight/templates/secret.yaml

@@ -1,9 +1,9 @@
 apiVersion: v1
 kind: Secret
 metadata:
-  name: {{ include "memora.fullname" . }}-secret
+  name: {{ include "hindsight.fullname" . }}-secret
   labels:
-    {{- include "memora.labels" . | nindent 4 }}
+    {{- include "hindsight.labels" . | nindent 4 }}
 type: Opaque
 data:
   {{- if and .Values.api.secrets (hasKey .Values.api.secrets "MEMORY_LLM_API_KEY") }}

helm/hindsight/templates/serviceaccount.yaml

@@ -2,9 +2,9 @@
 apiVersion: v1
 kind: ServiceAccount
 metadata:
-  name: {{ include "memora.serviceAccountName" . }}
+  name: {{ include "hindsight.serviceAccountName" . }}
   labels:
-    {{- include "memora.labels" . | nindent 4 }}
+    {{- include "hindsight.labels" . | nindent 4 }}
   {{- with .Values.serviceAccount.annotations }}
   annotations:
     {{- toYaml . | nindent 4 }}

helm/hindsight/values.yaml

@@ -1,4 +1,4 @@
-# Default values for memora
+# Default values for hindsight
 
 # Global settings
 replicaCount: 1
@@ -8,14 +8,14 @@ api:
   enabled: true
   replicaCount: 1
   image:
-    repository: memora/api
+    repository: hindsight/api
     pullPolicy: IfNotPresent
     tag: "latest"
 
   service:
     type: ClusterIP
-    port: 8080
-    targetPort: 8080
+    port: 8888
+    targetPort: 8888
 
   # Resource limits and requests
   resources:
@@ -30,7 +30,7 @@ api:
   livenessProbe:
     httpGet:
       path: /
-      port: 8080
+      port: 8888
     initialDelaySeconds: 30
     periodSeconds: 10
     timeoutSeconds: 5
@@ -39,7 +39,7 @@ api:
   readinessProbe:
     httpGet:
       path: /
-      port: 8080
+      port: 8888
     initialDelaySeconds: 10
     periodSeconds: 5
     timeoutSeconds: 3
@@ -47,20 +47,20 @@ api:
 
   # Environment variables
   env:
-    MEMORA_API_LLM_PROVIDER: "groq"
-    MEMORA_API_LLM_MODEL: "openai/gpt-oss-120b"
+    HINDSIGHT_API_LLM_PROVIDER: "groq"
+    HINDSIGHT_API_LLM_MODEL: "openai/gpt-oss-120b"
 
   # Secret environment variables
   secrets:
-    # MEMORA_API_LLM_API_KEY: "your-api-key"
-    # MEMORA_API_LLM_BASE_URL: "https://api.groq.com/openai/v1"
+    # HINDSIGHT_API_LLM_API_KEY: "your-api-key"
+    # HINDSIGHT_API_LLM_BASE_URL: "https://api.groq.com/openai/v1"
 
 # Image settings for control plane
 controlPlane:
   enabled: true
   replicaCount: 1
   image:
-    repository: memora/memora-control-plane
+    repository: hindsight/hindsight-control-plane
     pullPolicy: IfNotPresent
     tag: "latest"
 
@@ -100,8 +100,8 @@ controlPlane:
   # Environment variables
   env:
     NODE_ENV: "production"
-    MEMORA_CP_HOSTNAME: "0.0.0.0"
-    MEMORA_CP_PORT: "3000"
+    HINDSIGHT_CP_HOSTNAME: "0.0.0.0"
+    HINDSIGHT_CP_PORT: "3000"
 
 # PostgreSQL configuration
 postgresql:
@@ -113,8 +113,8 @@ postgresql:
   external:
     host: "postgresql"
     port: 5432
-    database: "memora"
-    username: "memora"
+    database: "hindsight"
+    username: "hindsight"
     # Password should be provided via secret
     # password: ""
 
@@ -130,7 +130,7 @@ ingress:
     # nginx.ingress.kubernetes.io/ssl-redirect: "true"
 
   hosts:
-    - host: memora.example.com
+    - host: hindsight.example.com
       paths:
         - path: /
           pathType: Prefix
@@ -140,9 +140,9 @@ ingress:
           service: api
 
   tls: []
-  #  - secretName: memora-tls
+  #  - secretName: hindsight-tls
   #    hosts:
-  #      - memora.example.com
+  #      - hindsight.example.com
 
 # Service Account
 serviceAccount:

helm/memora/Chart.yaml

@@ -1,13 +0,0 @@
-apiVersion: v2
-name: memora
-description: A Helm chart for Memora - temporal-semantic-entity memory system for AI agents
-type: application
-version: 0.0.7
-appVersion: "0.0.7"
-keywords:
-  - ai
-  - memory
-  - llm
-  - agents
-maintainers:
-  - name: Memora Team

helm/memora/templates/configmap.yaml

@@ -1,15 +0,0 @@
-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: {{ include "memora.fullname" . }}-config
-  labels:
-    {{- include "memora.labels" . | nindent 4 }}
-data:
-  # API configuration
-  llm-provider: {{ .Values.api.env.MEMORA_API_LLM_PROVIDER | quote }}
-  llm-model: {{ .Values.api.env.MEMORA_API_LLM_MODEL | quote }}
-
-  # Control plane configuration
-  node-env: {{ .Values.controlPlane.env.NODE_ENV | quote }}
-  hostname: {{ .Values.controlPlane.env.MEMORA_CP_HOSTNAME | quote }}
-  control-plane-port: {{ .Values.controlPlane.env.MEMORA_CP_PORT | quote }}

hindsight-api/alembic/env.py

@@ -14,13 +14,13 @@ from alembic import context
 from dotenv import load_dotenv
 
 # Import your models here
-from memora.models import Base
+from hindsight_api.models import Base
 
-# Load environment variables based on MEMORA_API_DATABASE_URL env var or default to local
+# Load environment variables based on HINDSIGHT_API_DATABASE_URL env var or default to local
 def load_env():
     """Load environment variables from .env"""
-    # Check if MEMORA_API_DATABASE_URL is already set (e.g., by CI/CD)
-    if os.getenv("MEMORA_API_DATABASE_URL"):
+    # Check if HINDSIGHT_API_DATABASE_URL is already set (e.g., by CI/CD)
+    if os.getenv("HINDSIGHT_API_DATABASE_URL"):
         return
 
     # Look for .env file in the parent directory (root of the workspace)
@@ -58,11 +58,11 @@ def get_database_url() -> str:
     # Get database URL from config (set programmatically) or environment
     database_url = config.get_main_option("sqlalchemy.url")
     if not database_url:
-        database_url = os.getenv("MEMORA_API_DATABASE_URL")
+        database_url = os.getenv("HINDSIGHT_API_DATABASE_URL")
         if not database_url:
             raise ValueError(
                 "Database URL not found. "
-                "Set MEMORA_API_DATABASE_URL environment variable or pass database_url to run_migrations()."
+                "Set HINDSIGHT_API_DATABASE_URL environment variable or pass database_url to run_migrations()."
             )
 
     # For migrations, use psycopg2 (sync driver) to avoid pgbouncer prepared statement issues

hindsight-api/hindsight_api/__init__.py

@@ -3,8 +3,8 @@ Memory System for AI Agents.
 
 Temporal + Semantic Memory Architecture using PostgreSQL with pgvector.
 """
-from .temporal_semantic_memory import TemporalSemanticMemory
-from .search_trace import (
+from .engine.memory_engine import MemoryEngine
+from .engine.search_trace import (
     SearchTrace,
     QueryInfo,
     EntryPoint,
@@ -15,11 +15,12 @@ from .search_trace import (
     SearchSummary,
     SearchPhaseMetrics,
 )
-from .search_tracer import SearchTracer
-from .embeddings import Embeddings, SentenceTransformersEmbeddings
+from .engine.search_tracer import SearchTracer
+from .engine.embeddings import Embeddings, SentenceTransformersEmbeddings
+from .engine.llm_wrapper import LLMConfig
 
 __all__ = [
-    "TemporalSemanticMemory",
+    "MemoryEngine",
     "SearchTrace",
     "SearchTracer",
     "QueryInfo",
@@ -32,5 +33,6 @@ __all__ = [
     "SearchPhaseMetrics",
     "Embeddings",
     "SentenceTransformersEmbeddings",
+    "LLMConfig",
 ]
 __version__ = "0.1.0"

hindsight-api/hindsight_api/api/__init__.py

@@ -1,5 +1,5 @@
 """
-Unified API module for Memora.
+Unified API module for Hindsight.
 
 Provides both HTTP REST API and MCP (Model Context Protocol) server.
 """
@@ -7,13 +7,13 @@ import logging
 from typing import Optional
 from fastapi import FastAPI
 
-from memora import TemporalSemanticMemory
+from hindsight_api import MemoryEngine
 
 logger = logging.getLogger(__name__)
 
 
 def create_app(
-    memory: TemporalSemanticMemory,
+    memory: MemoryEngine,
     http_api_enabled: bool = True,
     mcp_api_enabled: bool = False,
     mcp_mount_path: str = "/mcp",
@@ -21,10 +21,10 @@ def create_app(
     initialize_memory: bool = True
 ) -> FastAPI:
     """
-    Create and configure the unified Memora API application.
+    Create and configure the unified Hindsight API application.
 
     Args:
-        memory: TemporalSemanticMemory instance (already initialized with required parameters)
+        memory: MemoryEngine instance (already initialized with required parameters)
         http_api_enabled: Whether to enable HTTP REST API endpoints (default: True)
         mcp_api_enabled: Whether to enable MCP server (default: False)
         mcp_mount_path: Path to mount MCP server (default: /mcp)
@@ -56,7 +56,7 @@ def create_app(
         logger.info("HTTP REST API enabled")
     else:
         # Create minimal FastAPI app
-        app = FastAPI(title="Memora API", version="0.0.7")
+        app = FastAPI(title="Hindsight API", version="0.0.7")
         logger.info("HTTP REST API disabled")
 
     # Mount MCP server if enabled
@@ -67,12 +67,12 @@ def create_app(
             # Create MCP server with shared memory instance
             mcp_server = create_mcp_server(memory=memory)
 
-            # Mount at specified path
-            app.mount(mcp_mount_path, mcp_server.sse_app())
-            logger.info(f"MCP server enabled at {mcp_mount_path}/sse")
+            # Mount at specified path using http_app (modern non-SSE alternative)
+            app.mount(mcp_mount_path, mcp_server.http_app())
+            logger.info(f"MCP server enabled at {mcp_mount_path}")
         except ImportError as e:
             logger.error(f"MCP server requested but dependencies not available: {e}")
-            logger.error("Install with: pip install memora[mcp]")
+            logger.error("Install with: pip install hindsight-api[mcp]")
             raise
 
     return app
@@ -87,6 +87,8 @@ from .http import (
     BatchPutRequest,
     ThinkRequest,
     ThinkResponse,
+    CreateAgentRequest,
+    PersonalityTraits,
 )
 
 __all__ = [
@@ -98,4 +100,6 @@ __all__ = [
     "BatchPutRequest",
     "ThinkRequest",
     "ThinkResponse",
+    "CreateAgentRequest",
+    "PersonalityTraits",
 ]

hindsight-api/hindsight_api/api/http.py

@@ -14,29 +14,41 @@ from contextlib import asynccontextmanager
 from fastapi import FastAPI, HTTPException, Query
 from fastapi.staticfiles import StaticFiles
 from fastapi.responses import FileResponse
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, ConfigDict
 
-from memora import TemporalSemanticMemory
+from hindsight_api import MemoryEngine
+from hindsight_api.engine.db_utils import acquire_with_retry
 
 
 class MetadataFilter(BaseModel):
     """Filter for metadata fields. Matches records where (key=value) OR (key not set) when match_unset=True."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "key": "source",
+            "value": "slack",
+            "match_unset": True
+        }
+    })
+
     key: str = Field(description="Metadata key to filter on")
     value: Optional[str] = Field(default=None, description="Value to match. If None with match_unset=True, matches any record where key is not set.")
     match_unset: bool = Field(default=True, description="If True, also match records where this metadata key is not set")
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "key": "source",
-                "value": "slack",
-                "match_unset": True
-            }
-        }
-
 
 class SearchRequest(BaseModel):
     """Request model for search endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "query": "What did Alice say about machine learning?",
+            "fact_type": ["world", "agent"],
+            "thinking_budget": 100,
+            "max_tokens": 4096,
+            "trace": True,
+            "question_date": "2023-05-30T23:40:00",
+            "metadata_filter": [{"key": "source", "value": "slack", "match_unset": True}]
+        }
+    })
+
     query: str
     fact_type: Optional[List[str]] = None  # List of fact types to search (defaults to all if not specified)
     thinking_budget: int = 100
@@ -45,19 +57,6 @@ class SearchRequest(BaseModel):
     question_date: Optional[str] = None  # ISO format date string (e.g., "2023-05-30T23:40:00")
     metadata_filter: Optional[List[MetadataFilter]] = Field(default=None, description="Filter by metadata. Multiple filters are ANDed together.")
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "query": "What did Alice say about machine learning?",
-                "fact_type": ["world", "agent"],
-                "thinking_budget": 100,
-                "max_tokens": 4096,
-                "trace": True,
-                "question_date": "2023-05-30T23:40:00",
-                "metadata_filter": [{"key": "source", "value": "slack", "match_unset": True}]
-            }
-        }
-
 
 class SearchResult(BaseModel):
     """Single search result item."""
@@ -87,93 +86,100 @@ class SearchResult(BaseModel):
 
 class SearchResponse(BaseModel):
     """Response model for search endpoints."""
-    results: List[SearchResult]
-    trace: Optional[Dict[str, Any]] = None
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "results": [
-                    {
-                        "id": "123e4567-e89b-12d3-a456-426614174000",
-                        "text": "Alice works at Google on the AI team",
-                        "type": "world",
-                        "context": "work info",
-                        "event_date": "2024-01-15T10:30:00Z"
-                    }
-                ],
-                "trace": {
-                    "query": "What did Alice say about machine learning?",
-                    "num_results": 1,
-                    "time_seconds": 0.123
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "results": [
+                {
+                    "id": "123e4567-e89b-12d3-a456-426614174000",
+                    "text": "Alice works at Google on the AI team",
+                    "type": "world",
+                    "context": "work info",
+                    "event_date": "2024-01-15T10:30:00Z"
                 }
+            ],
+            "trace": {
+                "query": "What did Alice say about machine learning?",
+                "num_results": 1,
+                "time_seconds": 0.123
             }
         }
+    })
+
+    results: List[SearchResult]
+    trace: Optional[Dict[str, Any]] = None
 
 
 class MemoryItem(BaseModel):
     """Single memory item for batch put."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "content": "Alice mentioned she's working on a new ML model",
+            "event_date": "2024-01-15T10:30:00Z",
+            "context": "team meeting",
+            "metadata": {"source": "slack", "channel": "engineering"}
+        }
+    })
+
     content: str
     event_date: Optional[datetime] = None
     context: Optional[str] = None
     metadata: Optional[Dict[str, str]] = None
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "content": "Alice mentioned she's working on a new ML model",
-                "event_date": "2024-01-15T10:30:00Z",
-                "context": "team meeting",
-                "metadata": {"source": "slack", "channel": "engineering"}
-            }
-        }
-
 
 class BatchPutRequest(BaseModel):
     """Request model for batch put endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "items": [
+                {
+                    "content": "Alice works at Google",
+                    "context": "work"
+                },
+                {
+                    "content": "Bob went hiking yesterday",
+                    "event_date": "2024-01-15T10:00:00Z"
+                }
+            ],
+            "document_id": "conversation_123"
+        }
+    })
+
     items: List[MemoryItem]
     document_id: Optional[str] = None
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "items": [
-                    {
-                        "content": "Alice works at Google",
-                        "context": "work"
-                    },
-                    {
-                        "content": "Bob went hiking yesterday",
-                        "event_date": "2024-01-15T10:00:00Z"
-                    }
-                ],
-                "document_id": "conversation_123"
-            }
-        }
-
 
 class BatchPutResponse(BaseModel):
     """Response model for batch put endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "success": True,
+            "message": "Successfully stored 2 memory items",
+            "agent_id": "user123",
+            "document_id": "conversation_123",
+            "items_count": 2
+        }
+    })
+
     success: bool
     message: str
     agent_id: str
     document_id: Optional[str] = None
     items_count: int
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "success": True,
-                "message": "Successfully stored 2 memory items",
-                "agent_id": "user123",
-                "document_id": "conversation_123",
-                "items_count": 2
-            }
-        }
-
 
 class BatchPutAsyncResponse(BaseModel):
     """Response model for async batch put endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "success": True,
+            "message": "Batch put task queued for background processing",
+            "agent_id": "user123",
+            "document_id": "conversation_123",
+            "items_count": 2,
+            "queued": True
+        }
+    })
+
     success: bool
     message: str
     agent_id: str
@@ -181,36 +187,23 @@ class BatchPutAsyncResponse(BaseModel):
     items_count: int
     queued: bool
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "success": True,
-                "message": "Batch put task queued for background processing",
-                "agent_id": "user123",
-                "document_id": "conversation_123",
-                "items_count": 2,
-                "queued": True
-            }
-        }
-
 
 class ThinkRequest(BaseModel):
     """Request model for think endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "query": "What do you think about artificial intelligence?",
+            "thinking_budget": 50,
+            "context": "This is for a research paper on AI ethics",
+            "metadata_filter": [{"key": "source", "value": "slack", "match_unset": True}]
+        }
+    })
+
     query: str
     thinking_budget: int = 50
     context: Optional[str] = None
     metadata_filter: Optional[List[MetadataFilter]] = Field(default=None, description="Filter by metadata. Multiple filters are ANDed together.")
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "query": "What do you think about artificial intelligence?",
-                "thinking_budget": 50,
-                "context": "This is for a research paper on AI ethics",
-                "metadata_filter": [{"key": "source", "value": "slack", "match_unset": True}]
-            }
-        }
-
 
 class OpinionItem(BaseModel):
     """Model for an opinion with confidence score."""
@@ -220,67 +213,75 @@ class OpinionItem(BaseModel):
 
 class ThinkFact(BaseModel):
     """A fact used in think response."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "id": "123e4567-e89b-12d3-a456-426614174000",
+            "text": "AI is used in healthcare",
+            "type": "world",
+            "context": "healthcare discussion",
+            "event_date": "2024-01-15T10:30:00Z"
+        }
+    })
+
     id: Optional[str] = None
     text: str
     type: Optional[str] = None  # fact type: world, agent, opinion
     context: Optional[str] = None
     event_date: Optional[str] = None
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "id": "123e4567-e89b-12d3-a456-426614174000",
-                "text": "AI is used in healthcare",
-                "type": "world",
-                "context": "healthcare discussion",
-                "event_date": "2024-01-15T10:30:00Z"
-            }
-        }
-
 
 class ThinkResponse(BaseModel):
     """Response model for think endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "text": "Based on my understanding, AI is a transformative technology...",
+            "based_on": [
+                {
+                    "id": "123",
+                    "text": "AI is used in healthcare",
+                    "type": "world"
+                },
+                {
+                    "id": "456",
+                    "text": "I discussed AI applications last week",
+                    "type": "agent"
+                }
+            ],
+            "new_opinions": [
+                "AI has great potential when used responsibly"
+            ]
+        }
+    })
+
     text: str
     based_on: List[ThinkFact] = []  # Facts used to generate the response
     new_opinions: List[str] = []  # Simplified to list of opinion strings
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "text": "Based on my understanding, AI is a transformative technology...",
-                "based_on": [
-                    {
-                        "id": "123",
-                        "text": "AI is used in healthcare",
-                        "type": "world"
-                    },
-                    {
-                        "id": "456",
-                        "text": "I discussed AI applications last week",
-                        "type": "agent"
-                    }
-                ],
-                "new_opinions": [
-                    "AI has great potential when used responsibly"
-                ]
-            }
-        }
-
 
 class AgentsResponse(BaseModel):
     """Response model for agents list endpoint."""
-    agents: List[str]
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "agents": ["user123", "agent_alice", "agent_bob"]
-            }
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "agents": ["user123", "agent_alice", "agent_bob"]
         }
+    })
+
+    agents: List[str]
 
 
 class PersonalityTraits(BaseModel):
     """Personality traits based on Big Five model."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "openness": 0.8,
+            "conscientiousness": 0.6,
+            "extraversion": 0.5,
+            "agreeableness": 0.7,
+            "neuroticism": 0.3,
+            "bias_strength": 0.7
+        }
+    })
+
     openness: float = Field(ge=0.0, le=1.0, description="Openness to experience (0-1)")
     conscientiousness: float = Field(ge=0.0, le=1.0, description="Conscientiousness (0-1)")
     extraversion: float = Field(ge=0.0, le=1.0, description="Extraversion (0-1)")
@@ -288,43 +289,30 @@ class PersonalityTraits(BaseModel):
     neuroticism: float = Field(ge=0.0, le=1.0, description="Neuroticism (0-1)")
     bias_strength: float = Field(ge=0.0, le=1.0, description="How strongly personality influences opinions (0-1)")
 
-    class Config:
-        json_schema_extra = {
-            "example": {
+
+class AgentProfileResponse(BaseModel):
+    """Response model for agent profile."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "agent_id": "user123",
+            "name": "Alice",
+            "personality": {
                 "openness": 0.8,
                 "conscientiousness": 0.6,
                 "extraversion": 0.5,
                 "agreeableness": 0.7,
                 "neuroticism": 0.3,
                 "bias_strength": 0.7
-            }
+            },
+            "background": "I am a software engineer with 10 years of experience in startups"
         }
+    })
 
-
-class AgentProfileResponse(BaseModel):
-    """Response model for agent profile."""
     agent_id: str
     name: str
     personality: PersonalityTraits
     background: str
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "agent_id": "user123",
-                "name": "Alice",
-                "personality": {
-                    "openness": 0.8,
-                    "conscientiousness": 0.6,
-                    "extraversion": 0.5,
-                    "agreeableness": 0.7,
-                    "neuroticism": 0.3,
-                    "bias_strength": 0.7
-                },
-                "background": "I am a software engineer with 10 years of experience in startups"
-            }
-        }
-
 
 class UpdatePersonalityRequest(BaseModel):
     """Request model for updating personality traits."""
@@ -333,40 +321,38 @@ class UpdatePersonalityRequest(BaseModel):
 
 class AddBackgroundRequest(BaseModel):
     """Request model for adding/merging background information."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "content": "I was born in Texas",
+            "update_personality": True
+        }
+    })
+
     content: str = Field(description="New background information to add or merge")
     update_personality: bool = Field(
         default=True,
         description="If true, infer Big Five personality traits from the merged background (default: true)"
     )
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "content": "I was born in Texas",
-                "update_personality": True
-            }
-        }
-
 
 class BackgroundResponse(BaseModel):
     """Response model for background update."""
-    background: str
-    personality: Optional[PersonalityTraits] = None
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "background": "I was born in Texas. I am a software engineer with 10 years of experience.",
-                "personality": {
-                    "openness": 0.7,
-                    "conscientiousness": 0.6,
-                    "extraversion": 0.5,
-                    "agreeableness": 0.8,
-                    "neuroticism": 0.4,
-                    "bias_strength": 0.6
-                }
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "background": "I was born in Texas. I am a software engineer with 10 years of experience.",
+            "personality": {
+                "openness": 0.7,
+                "conscientiousness": 0.6,
+                "extraversion": 0.5,
+                "agreeableness": 0.8,
+                "neuroticism": 0.4,
+                "bias_strength": 0.6
             }
         }
+    })
+
+    background: str
+    personality: Optional[PersonalityTraits] = None
 
 
 class AgentListItem(BaseModel):
@@ -381,137 +367,144 @@ class AgentListItem(BaseModel):
 
 class AgentListResponse(BaseModel):
     """Response model for listing all agents."""
-    agents: List[AgentListItem]
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "agents": [
-                    {
-                        "agent_id": "user123",
-                        "name": "Alice",
-                        "personality": {
-                            "openness": 0.5,
-                            "conscientiousness": 0.5,
-                            "extraversion": 0.5,
-                            "agreeableness": 0.5,
-                            "neuroticism": 0.5,
-                            "bias_strength": 0.5
-                        },
-                        "background": "I am a software engineer",
-                        "created_at": "2024-01-15T10:30:00Z",
-                        "updated_at": "2024-01-16T14:20:00Z"
-                    }
-                ]
-            }
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "agents": [
+                {
+                    "agent_id": "user123",
+                    "name": "Alice",
+                    "personality": {
+                        "openness": 0.5,
+                        "conscientiousness": 0.5,
+                        "extraversion": 0.5,
+                        "agreeableness": 0.5,
+                        "neuroticism": 0.5,
+                        "bias_strength": 0.5
+                    },
+                    "background": "I am a software engineer",
+                    "created_at": "2024-01-15T10:30:00Z",
+                    "updated_at": "2024-01-16T14:20:00Z"
+                }
+            ]
         }
+    })
+
+    agents: List[AgentListItem]
 
 
 class CreateAgentRequest(BaseModel):
     """Request model for creating/updating an agent."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "name": "Alice",
+            "personality": {
+                "openness": 0.8,
+                "conscientiousness": 0.6,
+                "extraversion": 0.5,
+                "agreeableness": 0.7,
+                "neuroticism": 0.3,
+                "bias_strength": 0.7
+            },
+            "background": "I am a creative software engineer with 10 years of experience"
+        }
+    })
+
     name: Optional[str] = None
     personality: Optional[PersonalityTraits] = None
     background: Optional[str] = None
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "name": "Alice",
-                "personality": {
-                    "openness": 0.8,
-                    "conscientiousness": 0.6,
-                    "extraversion": 0.5,
-                    "agreeableness": 0.7,
-                    "neuroticism": 0.3,
-                    "bias_strength": 0.7
-                },
-                "background": "I am a creative software engineer with 10 years of experience"
-            }
-        }
-
 
 class GraphDataResponse(BaseModel):
     """Response model for graph data endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "nodes": [
+                {"id": "1", "label": "Alice works at Google", "type": "world"},
+                {"id": "2", "label": "Bob went hiking", "type": "world"}
+            ],
+            "edges": [
+                {"from": "1", "to": "2", "type": "semantic", "weight": 0.8}
+            ],
+            "table_rows": [
+                {"id": "abc12345...", "text": "Alice works at Google", "context": "Work info", "date": "2024-01-15 10:30", "entities": "Alice (PERSON), Google (ORGANIZATION)"}
+            ],
+            "total_units": 2
+        }
+    })
+
     nodes: List[Dict[str, Any]]
     edges: List[Dict[str, Any]]
     table_rows: List[Dict[str, Any]]
     total_units: int
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "nodes": [
-                    {"id": "1", "label": "Alice works at Google", "type": "world"},
-                    {"id": "2", "label": "Bob went hiking", "type": "world"}
-                ],
-                "edges": [
-                    {"from": "1", "to": "2", "type": "semantic", "weight": 0.8}
-                ],
-                "table_rows": [
-                    {"id": "abc12345...", "text": "Alice works at Google", "context": "Work info", "date": "2024-01-15 10:30", "entities": "Alice (PERSON), Google (ORGANIZATION)"}
-                ],
-                "total_units": 2
-            }
-        }
-
 
 class ListMemoryUnitsResponse(BaseModel):
     """Response model for list memory units endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "items": [
+                {
+                    "id": "550e8400-e29b-41d4-a716-446655440000",
+                    "text": "Alice works at Google on the AI team",
+                    "context": "Work conversation",
+                    "date": "2024-01-15T10:30:00Z",
+                    "fact_type": "world",
+                    "entities": "Alice (PERSON), Google (ORGANIZATION)"
+                }
+            ],
+            "total": 150,
+            "limit": 100,
+            "offset": 0
+        }
+    })
+
     items: List[Dict[str, Any]]
     total: int
     limit: int
     offset: int
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "items": [
-                    {
-                        "id": "550e8400-e29b-41d4-a716-446655440000",
-                        "text": "Alice works at Google on the AI team",
-                        "context": "Work conversation",
-                        "date": "2024-01-15T10:30:00Z",
-                        "fact_type": "world",
-                        "entities": "Alice (PERSON), Google (ORGANIZATION)"
-                    }
-                ],
-                "total": 150,
-                "limit": 100,
-                "offset": 0
-            }
-        }
-
 
 class ListDocumentsResponse(BaseModel):
     """Response model for list documents endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "items": [
+                {
+                    "id": "session_1",
+                    "agent_id": "user123",
+                    "content_hash": "abc123",
+                    "created_at": "2024-01-15T10:30:00Z",
+                    "updated_at": "2024-01-15T10:30:00Z",
+                    "text_length": 5420,
+                    "memory_unit_count": 15
+                }
+            ],
+            "total": 50,
+            "limit": 100,
+            "offset": 0
+        }
+    })
+
     items: List[Dict[str, Any]]
     total: int
     limit: int
     offset: int
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "items": [
-                    {
-                        "id": "session_1",
-                        "agent_id": "user123",
-                        "content_hash": "abc123",
-                        "created_at": "2024-01-15T10:30:00Z",
-                        "updated_at": "2024-01-15T10:30:00Z",
-                        "text_length": 5420,
-                        "memory_unit_count": 15
-                    }
-                ],
-                "total": 50,
-                "limit": 100,
-                "offset": 0
-            }
-        }
-
 
 class DocumentResponse(BaseModel):
     """Response model for get document endpoint."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "id": "session_1",
+            "agent_id": "user123",
+            "original_text": "Full document text here...",
+            "content_hash": "abc123",
+            "created_at": "2024-01-15T10:30:00Z",
+            "updated_at": "2024-01-15T10:30:00Z",
+            "memory_unit_count": 15
+        }
+    })
+
     id: str
     agent_id: str
     original_text: str
@@ -520,40 +513,26 @@ class DocumentResponse(BaseModel):
     updated_at: str
     memory_unit_count: int
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "id": "session_1",
-                "agent_id": "user123",
-                "original_text": "Full document text here...",
-                "content_hash": "abc123",
-                "created_at": "2024-01-15T10:30:00Z",
-                "updated_at": "2024-01-15T10:30:00Z",
-                "memory_unit_count": 15
-            }
-        }
-
 
 class DeleteResponse(BaseModel):
     """Response model for delete operations."""
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "success": True,
+            "message": "Resource deleted successfully"
+        }
+    })
+
     success: bool
     message: str
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "success": True,
-                "message": "Resource deleted successfully"
-            }
-        }
 
-
-def create_app(memory: TemporalSemanticMemory, run_migrations: bool = True, initialize_memory: bool = True) -> FastAPI:
+def create_app(memory: MemoryEngine, run_migrations: bool = True, initialize_memory: bool = True) -> FastAPI:
     """
     Create and configure the FastAPI application.
 
     Args:
-        memory: TemporalSemanticMemory instance (already initialized with required parameters)
+        memory: MemoryEngine instance (already initialized with required parameters)
         run_migrations: Whether to run database migrations on startup (default: True)
         initialize_memory: Whether to initialize memory system on startup (default: True)
 
@@ -572,15 +551,17 @@ def create_app(memory: TemporalSemanticMemory, run_migrations: bool = True, init
         Note: This only fires when running the app standalone, not when mounted.
         """
         # Startup: Initialize database and memory system
-        if run_migrations:
-            from memora.migrations import run_migrations as do_migrations
-            do_migrations(memory.db_url)
-            logging.info("Database migrations applied")
-
         if initialize_memory:
             await memory.initialize()
             logging.info("Memory system initialized")
 
+        if run_migrations:
+            from hindsight_api.migrations import run_migrations as do_migrations
+            do_migrations(memory.db_url)
+            logging.info("Database migrations applied")
+
+
+
         yield
 
         # Shutdown: Cleanup memory system
@@ -860,7 +841,7 @@ def _register_routes(app: FastAPI):
         """Get statistics about memory nodes and links for an agent."""
         try:
             pool = await app.state.memory._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 # Get node counts by fact_type
                 node_stats = await conn.fetch(
                     """
@@ -1195,7 +1176,7 @@ This operation cannot be undone.
 
             # Insert operation record into database BEFORE scheduling task
             pool = await app.state.memory._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 await conn.execute(
                     """
                     INSERT INTO async_operations (id, agent_id, task_type, items_count, document_id)
@@ -1245,7 +1226,7 @@ This operation cannot be undone.
         """List all async operations (pending and failed) for an agent."""
         try:
             pool = await app.state.memory._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 operations = await conn.fetch(
                     """
                     SELECT id, agent_id, task_type, items_count, document_id, created_at, status, error_message
@@ -1296,7 +1277,7 @@ This operation cannot be undone.
                 raise HTTPException(status_code=400, detail=f"Invalid operation_id format: {operation_id}")
 
             pool = await app.state.memory._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 # Check if operation exists and belongs to this agent
                 result = await conn.fetchrow(
                     "SELECT agent_id FROM async_operations WHERE id = $1 AND agent_id = $2",
@@ -1468,7 +1449,7 @@ This operation cannot be undone.
             # Update name if provided
             if request.name is not None:
                 pool = await app.state.memory._get_pool()
-                async with pool.acquire() as conn:
+                async with acquire_with_retry(pool) as conn:
                     await conn.execute(
                         """
                         UPDATE agents
@@ -1492,7 +1473,7 @@ This operation cannot be undone.
             # Update background if provided (replace, not merge)
             if request.background is not None:
                 pool = await app.state.memory._get_pool()
-                async with pool.acquire() as conn:
+                async with acquire_with_retry(pool) as conn:
                     await conn.execute(
                         """
                         UPDATE agents

hindsight-api/hindsight_api/api/mcp.py

@@ -1,30 +1,30 @@
-"""Memora MCP Server implementation using FastMCP."""
+"""Hindsight MCP Server implementation using FastMCP."""
 
 import json
 import logging
 
 from fastmcp import FastMCP
-from memora import TemporalSemanticMemory
+from hindsight_api import MemoryEngine
 
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
 
-def create_mcp_server(memory: TemporalSemanticMemory) -> FastMCP:
+def create_mcp_server(memory: MemoryEngine) -> FastMCP:
     """
-    Create and configure the Memora MCP server.
+    Create and configure the Hindsight MCP server.
 
     Args:
-        memory: TemporalSemanticMemory instance (required)
+        memory: MemoryEngine instance (required)
 
     Returns:
         Configured FastMCP server instance
     """
     # Create FastMCP server
-    mcp = FastMCP("memora-mcp-server")
+    mcp = FastMCP("hindsight-mcp-server")
 
     @mcp.tool()
-    async def memora_put(agent_id: str, content: str, context: str, explanation: str = "") -> str:
+    async def hindsight_put(agent_id: str, content: str, context: str, explanation: str = "") -> str:
         """
         **CRITICAL: Store important user information to long-term memory.**
 
@@ -74,7 +74,7 @@ def create_mcp_server(memory: TemporalSemanticMemory) -> FastMCP:
             return f"Error: {str(e)}"
 
     @mcp.tool()
-    async def memora_search(agent_id: str, query: str, max_tokens: int = 4096, explanation: str = "") -> str:
+    async def hindsight_search(agent_id: str, query: str, max_tokens: int = 4096, explanation: str = "") -> str:
         """
         **CRITICAL: Search user's memory to provide personalized, context-aware responses.**
 
@@ -112,7 +112,7 @@ def create_mcp_server(memory: TemporalSemanticMemory) -> FastMCP:
         """
         try:
             # Log all parameters for debugging
-            logger.info(f"memora_search called with: query={query!r}, max_tokens={max_tokens}, explanation={explanation!r}")
+            logger.info(f"hindsight_search called with: query={query!r}, max_tokens={max_tokens}, explanation={explanation!r}")
 
             # Log explanation if provided
             if explanation:

hindsight-api/hindsight_api/engine/__init__.py

@@ -0,0 +1,47 @@
+"""
+Memory Engine - Core implementation of the memory system.
+
+This package contains all the implementation details of the memory engine:
+- MemoryEngine: Main class for memory operations
+- Utility modules: embedding_utils, link_utils, think_utils, agent_utils
+- Supporting modules: embeddings, cross_encoder, entity_resolver, etc.
+"""
+
+from .memory_engine import MemoryEngine
+from .db_utils import acquire_with_retry
+from .embeddings import Embeddings, SentenceTransformersEmbeddings
+from .search_trace import (
+    SearchTrace,
+    QueryInfo,
+    EntryPoint,
+    NodeVisit,
+    WeightComponents,
+    LinkInfo,
+    PruningDecision,
+    SearchSummary,
+    SearchPhaseMetrics,
+)
+from .search_tracer import SearchTracer
+from .llm_wrapper import LLMConfig
+from .response_models import SearchResult, ThinkResult, MemoryFact
+
+__all__ = [
+    "MemoryEngine",
+    "acquire_with_retry",
+    "Embeddings",
+    "SentenceTransformersEmbeddings",
+    "SearchTrace",
+    "SearchTracer",
+    "QueryInfo",
+    "EntryPoint",
+    "NodeVisit",
+    "WeightComponents",
+    "LinkInfo",
+    "PruningDecision",
+    "SearchSummary",
+    "SearchPhaseMetrics",
+    "LLMConfig",
+    "SearchResult",
+    "ThinkResult",
+    "MemoryFact",
+]

hindsight-api/hindsight_api/engine/agent_utils.py

@@ -0,0 +1,426 @@
+"""
+Agent profile utilities for personality and background management.
+"""
+
+import json
+import logging
+import re
+from typing import Dict, Optional
+from pydantic import BaseModel, Field
+from .db_utils import acquire_with_retry
+
+logger = logging.getLogger(__name__)
+
+DEFAULT_PERSONALITY = {
+    "openness": 0.5,
+    "conscientiousness": 0.5,
+    "extraversion": 0.5,
+    "agreeableness": 0.5,
+    "neuroticism": 0.5,
+    "bias_strength": 0.5,
+}
+
+
+class PersonalityTraits(BaseModel):
+    """Big Five personality traits with bias strength (all values 0.0-1.0)."""
+    openness: float = Field(description="Creativity, curiosity, openness to new ideas (0.0-1.0)")
+    conscientiousness: float = Field(description="Organization, discipline, goal-directed (0.0-1.0)")
+    extraversion: float = Field(description="Sociability, assertiveness, energy from others (0.0-1.0)")
+    agreeableness: float = Field(description="Cooperation, empathy, consideration (0.0-1.0)")
+    neuroticism: float = Field(description="Emotional sensitivity, anxiety, stress response (0.0-1.0)")
+    bias_strength: float = Field(description="How much personality influences opinions (0.0-1.0)")
+
+
+class BackgroundMergeResponse(BaseModel):
+    """LLM response for background merge with personality inference."""
+    background: str = Field(description="Merged background in first person perspective")
+    personality: PersonalityTraits = Field(description="Inferred Big Five personality traits")
+
+
+async def get_agent_profile(pool, agent_id: str) -> Dict:
+    """
+    Get agent profile (name, personality + background).
+    Auto-creates agent with default values if not exists.
+
+    Args:
+        pool: Database connection pool
+        agent_id: Agent identifier
+
+    Returns:
+        Dict with 'name' (str), 'personality' (dict) and 'background' (str) keys
+    """
+    async with acquire_with_retry(pool) as conn:
+        # Try to get existing agent
+        row = await conn.fetchrow(
+            """
+            SELECT name, personality, background
+            FROM agents
+            WHERE agent_id = $1
+            """,
+            agent_id
+        )
+
+        if row:
+            # asyncpg returns JSONB as a string, so parse it
+            personality_data = row["personality"]
+            if isinstance(personality_data, str):
+                personality_data = json.loads(personality_data)
+
+            return {
+                "name": row["name"],
+                "personality": personality_data,
+                "background": row["background"]
+            }
+
+        # Agent doesn't exist, create with defaults
+        await conn.execute(
+            """
+            INSERT INTO agents (agent_id, name, personality, background)
+            VALUES ($1, $2, $3::jsonb, $4)
+            ON CONFLICT (agent_id) DO NOTHING
+            """,
+            agent_id,
+            agent_id,  # Default name is the agent_id
+            json.dumps(DEFAULT_PERSONALITY),
+            ""
+        )
+
+        return {
+            "name": agent_id,
+            "personality": DEFAULT_PERSONALITY.copy(),
+            "background": ""
+        }
+
+
+async def update_agent_personality(
+    pool,
+    agent_id: str,
+    personality: Dict[str, float]
+) -> None:
+    """
+    Update agent personality traits.
+
+    Args:
+        pool: Database connection pool
+        agent_id: Agent identifier
+        personality: Dict with Big Five traits + bias_strength (all 0-1)
+    """
+    # Ensure agent exists first
+    await get_agent_profile(pool, agent_id)
+
+    async with acquire_with_retry(pool) as conn:
+        await conn.execute(
+            """
+            UPDATE agents
+            SET personality = $2::jsonb,
+                updated_at = NOW()
+            WHERE agent_id = $1
+            """,
+            agent_id,
+            json.dumps(personality)
+        )
+
+
+async def merge_agent_background(
+    pool,
+    llm_config,
+    agent_id: str,
+    new_info: str,
+    update_personality: bool = True
+) -> dict:
+    """
+    Merge new background information with existing background using LLM.
+    Normalizes to first person ("I") and resolves conflicts.
+    Optionally infers personality traits from the merged background.
+
+    Args:
+        pool: Database connection pool
+        llm_config: LLM configuration for background merging
+        agent_id: Agent identifier
+        new_info: New background information to add/merge
+        update_personality: If True, infer Big Five traits from background (default: True)
+
+    Returns:
+        Dict with 'background' (str) and optionally 'personality' (dict) keys
+    """
+    # Get current profile
+    profile = await get_agent_profile(pool, agent_id)
+    current_background = profile["background"]
+
+    # Use LLM to merge backgrounds and optionally infer personality
+    result = await _llm_merge_background(
+        llm_config,
+        current_background,
+        new_info,
+        infer_personality=update_personality
+    )
+
+    merged_background = result["background"]
+    inferred_personality = result.get("personality")
+
+    # Update in database
+    async with acquire_with_retry(pool) as conn:
+        if inferred_personality:
+            # Update both background and personality
+            await conn.execute(
+                """
+                UPDATE agents
+                SET background = $2,
+                    personality = $3::jsonb,
+                    updated_at = NOW()
+                WHERE agent_id = $1
+                """,
+                agent_id,
+                merged_background,
+                json.dumps(inferred_personality)
+            )
+        else:
+            # Update only background
+            await conn.execute(
+                """
+                UPDATE agents
+                SET background = $2,
+                    updated_at = NOW()
+                WHERE agent_id = $1
+                """,
+                agent_id,
+                merged_background
+            )
+
+    response = {"background": merged_background}
+    if inferred_personality:
+        response["personality"] = inferred_personality
+
+    return response
+
+
+async def _llm_merge_background(
+    llm_config,
+    current: str,
+    new_info: str,
+    infer_personality: bool = False
+) -> dict:
+    """
+    Use LLM to intelligently merge background information.
+    Optionally infer Big Five personality traits from the merged background.
+
+    Args:
+        llm_config: LLM configuration to use
+        current: Current background text
+        new_info: New information to merge
+        infer_personality: If True, also infer personality traits
+
+    Returns:
+        Dict with 'background' (str) and optionally 'personality' (dict) keys
+    """
+    if infer_personality:
+        prompt = f"""You are helping maintain an agent's background/profile and infer their personality. You MUST respond with ONLY valid JSON.
+
+Current background: {current if current else "(empty)"}
+
+New information to add: {new_info}
+
+Instructions:
+1. Merge the new information with the current background
+2. If there are conflicts (e.g., different birthplaces), the NEW information overwrites the old
+3. Keep additions that don't conflict
+4. Output in FIRST PERSON ("I") perspective
+5. Be concise - keep merged background under 500 characters
+6. Infer Big Five personality traits from the merged background:
+   - Openness: 0.0-1.0 (creativity, curiosity, openness to new ideas)
+   - Conscientiousness: 0.0-1.0 (organization, discipline, goal-directed)
+   - Extraversion: 0.0-1.0 (sociability, assertiveness, energy from others)
+   - Agreeableness: 0.0-1.0 (cooperation, empathy, consideration)
+   - Neuroticism: 0.0-1.0 (emotional sensitivity, anxiety, stress response)
+   - Bias Strength: 0.0-1.0 (how much personality influences opinions)
+
+CRITICAL: You MUST respond with ONLY a valid JSON object. No markdown, no code blocks, no explanations. Just the JSON.
+
+Format:
+{{
+  "background": "the merged background text in first person",
+  "personality": {{
+    "openness": 0.7,
+    "conscientiousness": 0.6,
+    "extraversion": 0.5,
+    "agreeableness": 0.8,
+    "neuroticism": 0.4,
+    "bias_strength": 0.6
+  }}
+}}
+
+Trait inference examples:
+- "creative artist" → openness: 0.8+, bias_strength: 0.6
+- "organized engineer" → conscientiousness: 0.8+, openness: 0.5-0.6
+- "startup founder" → openness: 0.8+, extraversion: 0.7+, neuroticism: 0.3-0.4
+- "risk-averse analyst" → openness: 0.3-0.4, conscientiousness: 0.8+, neuroticism: 0.6+
+- "rational and diligent" → conscientiousness: 0.7+, openness: 0.6+
+- "passionate and dramatic" → extraversion: 0.7+, neuroticism: 0.6+, openness: 0.7+"""
+    else:
+        prompt = f"""You are helping maintain an agent's background/profile.
+
+Current background: {current if current else "(empty)"}
+
+New information to add: {new_info}
+
+Instructions:
+1. Merge the new information with the current background
+2. If there are conflicts (e.g., different birthplaces), the NEW information overwrites the old
+3. Keep additions that don't conflict
+4. Output in FIRST PERSON ("I") perspective
+5. Be concise - keep it under 500 characters
+6. Return ONLY the merged background text, no explanations
+
+Merged background:"""
+
+    try:
+        # Prepare messages
+        messages = [{"role": "user", "content": prompt}]
+
+        if infer_personality:
+            # Use structured output with Pydantic model for personality inference
+            try:
+                parsed = await llm_config.call(
+                    messages=messages,
+                    response_format=BackgroundMergeResponse,
+                    scope="agent_background",
+                    temperature=0.3,
+                    max_tokens=8192
+                )
+                logger.info(f"Successfully got structured response: background={parsed.background[:100]}")
+
+                # Convert Pydantic model to dict format
+                return {
+                    "background": parsed.background,
+                    "personality": parsed.personality.model_dump()
+                }
+            except Exception as e:
+                logger.warning(f"Structured output failed, falling back to manual parsing: {e}")
+                # Fall through to manual parsing below
+
+        # Manual parsing fallback or non-personality merge
+        content = await llm_config.call(
+            messages=messages,
+            scope="agent_background",
+            temperature=0.3,
+            max_tokens=8192
+        )
+
+        logger.info(f"LLM response for background merge (first 500 chars): {content[:500]}")
+
+        if infer_personality:
+            # Parse JSON response - try multiple extraction methods
+            result = None
+
+            # Method 1: Direct parse
+            try:
+                result = json.loads(content)
+                logger.info("Successfully parsed JSON directly")
+            except json.JSONDecodeError:
+                pass
+
+            # Method 2: Extract from markdown code blocks
+            if result is None:
+                # Remove markdown code blocks
+                code_block_match = re.search(r'```(?:json)?\s*(\{.*?\})\s*```', content, re.DOTALL)
+                if code_block_match:
+                    try:
+                        result = json.loads(code_block_match.group(1))
+                        logger.info("Successfully extracted JSON from markdown code block")
+                    except json.JSONDecodeError:
+                        pass
+
+            # Method 3: Find nested JSON structure
+            if result is None:
+                # Look for JSON object with nested structure
+                json_match = re.search(r'\{[^{}]*"background"[^{}]*"personality"[^{}]*\{[^{}]*\}[^{}]*\}', content, re.DOTALL)
+                if json_match:
+                    try:
+                        result = json.loads(json_match.group())
+                        logger.info("Successfully extracted JSON using nested pattern")
+                    except json.JSONDecodeError:
+                        pass
+
+            # All parsing methods failed - use fallback
+            if result is None:
+                logger.warning(f"Failed to extract JSON from LLM response. Raw content: {content[:200]}")
+                # Fallback: use new_info as background with default personality
+                return {
+                    "background": new_info if new_info else current if current else "",
+                    "personality": DEFAULT_PERSONALITY.copy()
+                }
+
+            # Validate personality values
+            personality = result.get("personality", {})
+            for key in ["openness", "conscientiousness", "extraversion",
+                       "agreeableness", "neuroticism", "bias_strength"]:
+                if key not in personality:
+                    personality[key] = 0.5  # Default to neutral
+                else:
+                    # Clamp to [0, 1]
+                    personality[key] = max(0.0, min(1.0, float(personality[key])))
+
+            result["personality"] = personality
+
+            # Ensure background exists
+            if "background" not in result or not result["background"]:
+                result["background"] = new_info if new_info else ""
+
+            return result
+        else:
+            # Just background merge
+            merged = content
+            if not merged or merged.lower() in ["(empty)", "none", "n/a"]:
+                merged = new_info if new_info else ""
+            return {"background": merged}
+
+    except Exception as e:
+        logger.error(f"Error merging background with LLM: {e}")
+        # Fallback: just append new info
+        if current:
+            merged = f"{current} {new_info}".strip()
+        else:
+            merged = new_info
+
+        result = {"background": merged}
+        if infer_personality:
+            result["personality"] = DEFAULT_PERSONALITY.copy()
+        return result
+
+
+async def list_agents(pool) -> list:
+    """
+    List all agents in the system.
+
+    Args:
+        pool: Database connection pool
+
+    Returns:
+        List of dicts with agent_id, name, personality, background, created_at, updated_at
+    """
+    async with acquire_with_retry(pool) as conn:
+        rows = await conn.fetch(
+            """
+            SELECT agent_id, name, personality, background, created_at, updated_at
+            FROM agents
+            ORDER BY updated_at DESC
+            """
+        )
+
+        result = []
+        for row in rows:
+            # asyncpg returns JSONB as a string, so parse it
+            personality_data = row["personality"]
+            if isinstance(personality_data, str):
+                personality_data = json.loads(personality_data)
+
+            result.append({
+                "agent_id": row["agent_id"],
+                "name": row["name"],
+                "personality": personality_data,
+                "background": row["background"],
+                "created_at": row["created_at"].isoformat() if row["created_at"] else None,
+                "updated_at": row["updated_at"].isoformat() if row["updated_at"] else None,
+            })
+
+        return result

hindsight-api/hindsight_api/engine/db_utils.py

@@ -0,0 +1,93 @@
+"""
+Database utility functions for connection management with retry logic.
+"""
+import asyncio
+import logging
+from contextlib import asynccontextmanager
+import asyncpg
+
+logger = logging.getLogger(__name__)
+
+# Default retry configuration for database operations
+DEFAULT_MAX_RETRIES = 3
+DEFAULT_BASE_DELAY = 0.5  # seconds
+DEFAULT_MAX_DELAY = 5.0  # seconds
+
+# Exceptions that indicate transient connection issues worth retrying
+RETRYABLE_EXCEPTIONS = (
+    asyncpg.exceptions.InterfaceError,
+    asyncpg.exceptions.ConnectionDoesNotExistError,
+    asyncpg.exceptions.TooManyConnectionsError,
+    OSError,
+    ConnectionError,
+    asyncio.TimeoutError,
+)
+
+
+async def retry_with_backoff(
+    func,
+    max_retries: int = DEFAULT_MAX_RETRIES,
+    base_delay: float = DEFAULT_BASE_DELAY,
+    max_delay: float = DEFAULT_MAX_DELAY,
+    retryable_exceptions: tuple = RETRYABLE_EXCEPTIONS,
+):
+    """
+    Execute an async function with exponential backoff retry.
+
+    Args:
+        func: Async function to execute
+        max_retries: Maximum number of retry attempts
+        base_delay: Initial delay between retries (seconds)
+        max_delay: Maximum delay between retries (seconds)
+        retryable_exceptions: Tuple of exception types to retry on
+
+    Returns:
+        Result of the function
+
+    Raises:
+        The last exception if all retries fail
+    """
+    last_exception = None
+    for attempt in range(max_retries + 1):
+        try:
+            return await func()
+        except retryable_exceptions as e:
+            last_exception = e
+            if attempt < max_retries:
+                delay = min(base_delay * (2 ** attempt), max_delay)
+                logger.warning(
+                    f"Database operation failed (attempt {attempt + 1}/{max_retries + 1}): {e}. "
+                    f"Retrying in {delay:.1f}s..."
+                )
+                await asyncio.sleep(delay)
+            else:
+                logger.error(
+                    f"Database operation failed after {max_retries + 1} attempts: {e}"
+                )
+    raise last_exception
+
+
+@asynccontextmanager
+async def acquire_with_retry(pool: asyncpg.Pool, max_retries: int = DEFAULT_MAX_RETRIES):
+    """
+    Async context manager to acquire a connection with retry logic.
+
+    Usage:
+        async with acquire_with_retry(pool) as conn:
+            await conn.execute(...)
+
+    Args:
+        pool: The asyncpg connection pool
+        max_retries: Maximum number of retry attempts
+
+    Yields:
+        An asyncpg connection
+    """
+    async def acquire():
+        return await pool.acquire()
+
+    conn = await retry_with_backoff(acquire, max_retries=max_retries)
+    try:
+        yield conn
+    finally:
+        await pool.release(conn)

hindsight-api/hindsight_api/engine/embedding_utils.py

@@ -0,0 +1,54 @@
+"""
+Embedding generation utilities for memory units.
+"""
+
+import asyncio
+import logging
+from typing import List
+
+logger = logging.getLogger(__name__)
+
+
+def generate_embedding(embeddings_backend, text: str) -> List[float]:
+    """
+    Generate embedding for text using the provided embeddings backend.
+
+    Args:
+        embeddings_backend: Embeddings instance to use for encoding
+        text: Text to embed
+
+    Returns:
+        Embedding vector (dimension depends on embeddings backend)
+    """
+    try:
+        embeddings = embeddings_backend.encode([text])
+        return embeddings[0]
+    except Exception as e:
+        raise Exception(f"Failed to generate embedding: {str(e)}")
+
+
+async def generate_embeddings_batch(embeddings_backend, texts: List[str]) -> List[List[float]]:
+    """
+    Generate embeddings for multiple texts using the provided embeddings backend.
+
+    Runs the embedding generation in a thread pool to avoid blocking the event loop
+    for CPU-bound operations.
+
+    Args:
+        embeddings_backend: Embeddings instance to use for encoding
+        texts: List of texts to embed
+
+    Returns:
+        List of embeddings in same order as input texts
+    """
+    try:
+        # Run embeddings in thread pool to avoid blocking event loop
+        loop = asyncio.get_event_loop()
+        embeddings = await loop.run_in_executor(
+            None,  # Use default thread pool
+            embeddings_backend.encode,
+            texts
+        )
+        return embeddings
+    except Exception as e:
+        raise Exception(f"Failed to generate batch embeddings: {str(e)}")

hindsight-api/hindsight_api/engine/entity_resolver.py

@@ -8,6 +8,7 @@ import asyncpg
 from typing import List, Dict, Optional, Set
 from difflib import SequenceMatcher
 from datetime import datetime, timezone
+from .db_utils import acquire_with_retry
 
 
 # Load spaCy model (singleton)
@@ -56,7 +57,7 @@ class EntityResolver:
             return []
 
         if conn is None:
-            async with self.pool.acquire() as conn:
+            async with acquire_with_retry(self.pool) as conn:
                 return await self._resolve_entities_batch_impl(conn, agent_id, entities_data, context, unit_event_date)
         else:
             return await self._resolve_entities_batch_impl(conn, agent_id, entities_data, context, unit_event_date)
@@ -220,7 +221,7 @@ class EntityResolver:
         Returns:
             Entity ID (creates new entity if needed)
         """
-        async with self.pool.acquire() as conn:
+        async with acquire_with_retry(self.pool) as conn:
             # Find candidate entities with similar name
             candidates = await conn.fetch(
                 """
@@ -366,7 +367,7 @@ class EntityResolver:
             unit_id: Memory unit ID
             entity_id: Entity ID
         """
-        async with self.pool.acquire() as conn:
+        async with acquire_with_retry(self.pool) as conn:
             # Insert unit-entity link
             await conn.execute(
                 """
@@ -434,7 +435,7 @@ class EntityResolver:
             return
 
         if conn is None:
-            async with self.pool.acquire() as conn:
+            async with acquire_with_retry(self.pool) as conn:
                 return await self._link_units_to_entities_batch_impl(conn, unit_entity_pairs)
         else:
             return await self._link_units_to_entities_batch_impl(conn, unit_entity_pairs)
@@ -499,7 +500,7 @@ class EntityResolver:
         Returns:
             List of unit IDs
         """
-        async with self.pool.acquire() as conn:
+        async with acquire_with_retry(self.pool) as conn:
             rows = await conn.fetch(
                 """
                 SELECT unit_id
@@ -527,7 +528,7 @@ class EntityResolver:
         Returns:
             Entity ID if found, None otherwise
         """
-        async with self.pool.acquire() as conn:
+        async with acquire_with_retry(self.pool) as conn:
             row = await conn.fetchrow(
                 """
                 SELECT id FROM entities

hindsight-api/hindsight_api/engine/link_utils.py

@@ -0,0 +1,541 @@
+"""
+Link creation utilities for temporal, semantic, and entity links.
+"""
+
+import time
+import logging
+from typing import List
+from datetime import timedelta
+
+logger = logging.getLogger(__name__)
+
+
+def _log(log_buffer, message, level='info'):
+    """Helper to log to buffer if available, otherwise use logger."""
+    if log_buffer is not None:
+        log_buffer.append(message)
+    else:
+        if level == 'info':
+            logger.info(message)
+        else:
+            logger.debug(message)
+
+
+async def extract_entities_batch_optimized(
+    entity_resolver,
+    conn,
+    agent_id: str,
+    unit_ids: List[str],
+    sentences: List[str],
+    context: str,
+    fact_dates: List,
+    llm_entities: List[List[dict]],
+    log_buffer: List[str] = None,
+) -> List[tuple]:
+    """
+    Process LLM-extracted entities for ALL facts in batch.
+
+    Uses entities provided by the LLM (no spaCy needed), then resolves
+    and links them in bulk.
+
+    Args:
+        entity_resolver: EntityResolver instance for entity resolution
+        conn: Database connection
+        agent_id: Agent identifier
+        unit_ids: List of unit IDs
+        sentences: List of fact sentences
+        context: Context string
+        fact_dates: List of fact dates
+        llm_entities: List of entity lists from LLM extraction
+        log_buffer: Optional buffer for logging
+
+    Returns:
+        List of tuples for batch insertion: (from_unit_id, to_unit_id, link_type, weight, entity_id)
+    """
+    try:
+        # Step 1: Convert LLM entities to the format expected by entity resolver
+        substep_start = time.time()
+        all_entities = []
+        for entity_list in llm_entities:
+            # Convert List[Entity] or List[dict] to List[Dict] format
+            formatted_entities = []
+            for ent in entity_list:
+                # Handle both Entity objects and dicts
+                if hasattr(ent, 'text'):
+                    formatted_entities.append({'text': ent.text, 'type': ent.type})
+                elif isinstance(ent, dict):
+                    formatted_entities.append({'text': ent.get('text', ''), 'type': ent.get('type', 'CONCEPT')})
+            all_entities.append(formatted_entities)
+
+        total_entities = sum(len(ents) for ents in all_entities)
+        _log(log_buffer, f"  [6.1] Process LLM entities: {total_entities} entities from {len(sentences)} facts in {time.time() - substep_start:.3f}s")
+
+        # Step 2: Resolve entities in BATCH (much faster!)
+        substep_start = time.time()
+        step_6_2_start = time.time()
+
+        # [6.2.1] Prepare all entities for batch resolution
+        substep_6_2_1_start = time.time()
+        all_entities_flat = []
+        entity_to_unit = []  # Maps flat index to (unit_id, local_index)
+
+        for unit_id, entities, fact_date in zip(unit_ids, all_entities, fact_dates):
+            if not entities:
+                continue
+
+            for local_idx, entity in enumerate(entities):
+                all_entities_flat.append({
+                    'text': entity['text'],
+                    'type': entity['type'],
+                    'nearby_entities': entities,
+                })
+                entity_to_unit.append((unit_id, local_idx, fact_date))
+        _log(log_buffer, f"    [6.2.1] Prepare entities: {len(all_entities_flat)} entities in {time.time() - substep_6_2_1_start:.3f}s")
+
+        # Resolve ALL entities in one batch call
+        if all_entities_flat:
+            # [6.2.2] Batch resolve entities
+            substep_6_2_2_start = time.time()
+            # Group by date for batch resolution (most will have same date)
+            entities_by_date = {}
+            for idx, (unit_id, local_idx, fact_date) in enumerate(entity_to_unit):
+                date_key = fact_date
+                if date_key not in entities_by_date:
+                    entities_by_date[date_key] = []
+                entities_by_date[date_key].append((idx, all_entities_flat[idx]))
+
+            _log(log_buffer, f"    [6.2.2] Grouped into {len(entities_by_date)} date buckets, resolving...")
+
+            # Resolve each date group in batch
+            resolved_entity_ids = [None] * len(all_entities_flat)
+            for date_idx, (fact_date, entities_group) in enumerate(entities_by_date.items(), 1):
+                date_bucket_start = time.time()
+                indices = [idx for idx, _ in entities_group]
+                entities_data = [entity_data for _, entity_data in entities_group]
+
+                batch_resolved = await entity_resolver.resolve_entities_batch(
+                    agent_id=agent_id,
+                    entities_data=entities_data,
+                    context=context,
+                    unit_event_date=fact_date,
+                    conn=conn
+                )
+
+                for idx, entity_id in zip(indices, batch_resolved):
+                    resolved_entity_ids[idx] = entity_id
+
+                _log(log_buffer, f"      [6.2.2.{date_idx}] Resolved {len(entities_data)} entities in {time.time() - date_bucket_start:.3f}s")
+
+            _log(log_buffer, f"    [6.2.2] Resolve entities: {len(all_entities_flat)} entities in {time.time() - substep_6_2_2_start:.3f}s")
+
+            # [6.2.3] Create unit-entity links in BATCH
+            substep_6_2_3_start = time.time()
+            # Map resolved entities back to units and collect all (unit, entity) pairs
+            unit_to_entity_ids = {}
+            unit_entity_pairs = []
+            for idx, (unit_id, local_idx, fact_date) in enumerate(entity_to_unit):
+                if unit_id not in unit_to_entity_ids:
+                    unit_to_entity_ids[unit_id] = []
+
+                entity_id = resolved_entity_ids[idx]
+                unit_to_entity_ids[unit_id].append(entity_id)
+                unit_entity_pairs.append((unit_id, entity_id))
+
+            # Batch insert all unit-entity links (MUCH faster!)
+            await entity_resolver.link_units_to_entities_batch(unit_entity_pairs, conn=conn)
+            _log(log_buffer, f"    [6.2.3] Create unit-entity links (batched): {len(unit_entity_pairs)} links in {time.time() - substep_6_2_3_start:.3f}s")
+
+            _log(log_buffer, f"  [6.2] Entity resolution (batched): {len(all_entities_flat)} entities resolved in {time.time() - step_6_2_start:.3f}s")
+        else:
+            unit_to_entity_ids = {}
+            _log(log_buffer, f"  [6.2] Entity resolution (batched): 0 entities in {time.time() - step_6_2_start:.3f}s")
+
+        # Step 3: Create entity links between units that share entities
+        substep_start = time.time()
+        # Collect all unique entity IDs
+        all_entity_ids = set()
+        for entity_ids in unit_to_entity_ids.values():
+            all_entity_ids.update(entity_ids)
+
+        _log(log_buffer, f"  [6.3] Creating entity links for {len(all_entity_ids)} unique entities...")
+
+        # Find all units that reference these entities (ONE batched query)
+        entity_to_units = {}
+        if all_entity_ids:
+            query_start = time.time()
+            import uuid
+            entity_id_list = [uuid.UUID(eid) if isinstance(eid, str) else eid for eid in all_entity_ids]
+            rows = await conn.fetch(
+                """
+                SELECT entity_id, unit_id
+                FROM unit_entities
+                WHERE entity_id = ANY($1::uuid[])
+                """,
+                entity_id_list
+            )
+            _log(log_buffer, f"      [6.3.1] Query unit_entities: {len(rows)} rows in {time.time() - query_start:.3f}s")
+
+            # Group by entity_id
+            group_start = time.time()
+            for row in rows:
+                entity_id = row['entity_id']
+                if entity_id not in entity_to_units:
+                    entity_to_units[entity_id] = []
+                entity_to_units[entity_id].append(row['unit_id'])
+            _log(log_buffer, f"      [6.3.2] Group by entity_id: {time.time() - group_start:.3f}s")
+
+        # Create bidirectional links between units that share entities
+        link_gen_start = time.time()
+        links = []
+        for entity_id, units_with_entity in entity_to_units.items():
+            # For each pair of units with this entity, create bidirectional links
+            for i, unit_id_1 in enumerate(units_with_entity):
+                for unit_id_2 in units_with_entity[i+1:]:
+                    # Bidirectional links
+                    links.append((unit_id_1, unit_id_2, 'entity', 1.0, entity_id))
+                    links.append((unit_id_2, unit_id_1, 'entity', 1.0, entity_id))
+
+        _log(log_buffer, f"      [6.3.3] Generate {len(links)} links: {time.time() - link_gen_start:.3f}s")
+        _log(log_buffer, f"  [6.3] Entity link creation: {len(links)} links for {len(all_entity_ids)} unique entities in {time.time() - substep_start:.3f}s")
+
+        return links
+
+    except Exception as e:
+        logger.error(f"Failed to extract entities in batch: {str(e)}")
+        import traceback
+        traceback.print_exc()
+        raise
+
+
+async def create_temporal_links_batch_per_fact(
+    conn,
+    agent_id: str,
+    unit_ids: List[str],
+    time_window_hours: int = 24,
+    log_buffer: List[str] = None,
+):
+    """
+    Create temporal links for multiple units, each with their own event_date.
+
+    Queries the event_date for each unit from the database and creates temporal
+    links based on individual dates (supports per-fact dating).
+
+    Args:
+        conn: Database connection
+        agent_id: Agent identifier
+        unit_ids: List of unit IDs
+        time_window_hours: Time window in hours for temporal links
+        log_buffer: Optional buffer for logging
+    """
+    if not unit_ids:
+        return
+
+    try:
+        import time as time_mod
+
+        # Get the event_date for each new unit
+        fetch_dates_start = time_mod.time()
+        rows = await conn.fetch(
+            """
+            SELECT id, event_date
+            FROM memory_units
+            WHERE id::text = ANY($1)
+            """,
+            unit_ids
+        )
+        new_units = {str(row['id']): row['event_date'] for row in rows}
+        _log(log_buffer, f"      [7.1] Fetch event_dates for {len(unit_ids)} units: {time_mod.time() - fetch_dates_start:.3f}s")
+
+        # Fetch ALL potential temporal neighbors in ONE query (much faster!)
+        # Get time range across all units
+        all_dates = list(new_units.values())
+        min_date = min(all_dates) - timedelta(hours=time_window_hours)
+        max_date = max(all_dates) + timedelta(hours=time_window_hours)
+
+        fetch_neighbors_start = time_mod.time()
+        all_candidates = await conn.fetch(
+            """
+            SELECT id, event_date
+            FROM memory_units
+            WHERE agent_id = $1
+              AND event_date BETWEEN $2 AND $3
+              AND id::text != ALL($4)
+            ORDER BY event_date DESC
+            """,
+            agent_id,
+            min_date,
+            max_date,
+            unit_ids
+        )
+        _log(log_buffer, f"      [7.2] Fetch {len(all_candidates)} candidate neighbors (1 query): {time_mod.time() - fetch_neighbors_start:.3f}s")
+
+        # Filter and create links in memory (much faster than N queries)
+        link_gen_start = time_mod.time()
+        links = []
+        for unit_id, unit_event_date in new_units.items():
+            # Filter candidates within this unit's time window
+            time_lower = unit_event_date - timedelta(hours=time_window_hours)
+            time_upper = unit_event_date + timedelta(hours=time_window_hours)
+
+            matching_neighbors = [
+                (row['id'], row['event_date'])
+                for row in all_candidates
+                if time_lower <= row['event_date'] <= time_upper
+            ][:10]  # Limit to top 10
+
+            for recent_id, recent_event_date in matching_neighbors:
+                # Calculate temporal proximity weight
+                time_diff_hours = abs((unit_event_date - recent_event_date).total_seconds() / 3600)
+                weight = max(0.3, 1.0 - (time_diff_hours / time_window_hours))
+                links.append((unit_id, str(recent_id), 'temporal', weight, None))
+
+        _log(log_buffer, f"      [7.3] Generate {len(links)} temporal links: {time_mod.time() - link_gen_start:.3f}s")
+
+        if links:
+            insert_start = time_mod.time()
+            await conn.executemany(
+                """
+                INSERT INTO memory_links (from_unit_id, to_unit_id, link_type, weight, entity_id)
+                VALUES ($1, $2, $3, $4, $5)
+                ON CONFLICT (from_unit_id, to_unit_id, link_type, COALESCE(entity_id, '00000000-0000-0000-0000-000000000000'::uuid)) DO NOTHING
+                """,
+                links
+            )
+            _log(log_buffer, f"      [7.4] Insert {len(links)} temporal links: {time_mod.time() - insert_start:.3f}s")
+
+    except Exception as e:
+        logger.error(f"Failed to create temporal links: {str(e)}")
+        import traceback
+        traceback.print_exc()
+        raise
+
+
+async def create_semantic_links_batch(
+    conn,
+    agent_id: str,
+    unit_ids: List[str],
+    embeddings: List[List[float]],
+    top_k: int = 5,
+    threshold: float = 0.7,
+    log_buffer: List[str] = None,
+):
+    """
+    Create semantic links for multiple units efficiently.
+
+    For each unit, finds similar units and creates links.
+
+    Args:
+        conn: Database connection
+        agent_id: Agent identifier
+        unit_ids: List of unit IDs
+        embeddings: List of embedding vectors
+        top_k: Number of top similar units to link
+        threshold: Minimum similarity threshold
+        log_buffer: Optional buffer for logging
+    """
+    if not unit_ids or not embeddings:
+        return
+
+    try:
+        import time as time_mod
+        import numpy as np
+
+        # Fetch ALL existing units with embeddings in ONE query
+        fetch_start = time_mod.time()
+        all_existing = await conn.fetch(
+            """
+            SELECT id, embedding
+            FROM memory_units
+            WHERE agent_id = $1
+              AND embedding IS NOT NULL
+              AND id::text != ALL($2)
+            """,
+            agent_id,
+            unit_ids
+        )
+        _log(log_buffer, f"      [8.1] Fetch {len(all_existing)} existing embeddings (1 query): {time_mod.time() - fetch_start:.3f}s")
+
+        # Convert to numpy for vectorized similarity computation
+        compute_start = time_mod.time()
+        all_links = []
+
+        if all_existing:
+            # Convert existing embeddings to numpy array
+            existing_ids = [str(row['id']) for row in all_existing]
+            # Stack embeddings as 2D array: (num_embeddings, embedding_dim)
+            embedding_arrays = []
+            for row in all_existing:
+                raw_emb = row['embedding']
+                # Handle different pgvector formats
+                if isinstance(raw_emb, str):
+                    # Parse string format: "[1.0, 2.0, ...]"
+                    import json
+                    emb = np.array(json.loads(raw_emb), dtype=np.float32)
+                elif isinstance(raw_emb, (list, tuple)):
+                    emb = np.array(raw_emb, dtype=np.float32)
+                else:
+                    # Try direct conversion (works for numpy arrays, pgvector objects, etc.)
+                    emb = np.array(raw_emb, dtype=np.float32)
+
+                # Ensure it's 1D
+                if emb.ndim != 1:
+                    raise ValueError(f"Expected 1D embedding, got shape {emb.shape}")
+                embedding_arrays.append(emb)
+
+            if not embedding_arrays:
+                existing_embeddings = np.array([])
+            elif len(embedding_arrays) == 1:
+                # Single embedding: reshape to (1, dim)
+                existing_embeddings = embedding_arrays[0].reshape(1, -1)
+            else:
+                # Multiple embeddings: vstack
+                existing_embeddings = np.vstack(embedding_arrays)
+
+            # For each new unit, compute similarities with ALL existing units
+            for unit_id, new_embedding in zip(unit_ids, embeddings):
+                new_emb_array = np.array(new_embedding)
+
+                # Compute cosine similarities (dot product for normalized vectors)
+                similarities = np.dot(existing_embeddings, new_emb_array)
+
+                # Find top-k above threshold
+                # Get indices of similarities above threshold
+                above_threshold = np.where(similarities >= threshold)[0]
+
+                if len(above_threshold) > 0:
+                    # Sort by similarity (descending) and take top-k
+                    sorted_indices = above_threshold[np.argsort(-similarities[above_threshold])][:top_k]
+
+                    for idx in sorted_indices:
+                        similar_id = existing_ids[idx]
+                        similarity = float(similarities[idx])
+                        all_links.append((unit_id, similar_id, 'semantic', similarity, None))
+
+        _log(log_buffer, f"      [8.2] Compute similarities & generate {len(all_links)} semantic links: {time_mod.time() - compute_start:.3f}s")
+
+        if all_links:
+            insert_start = time_mod.time()
+            await conn.executemany(
+                """
+                INSERT INTO memory_links (from_unit_id, to_unit_id, link_type, weight, entity_id)
+                VALUES ($1, $2, $3, $4, $5)
+                ON CONFLICT (from_unit_id, to_unit_id, link_type, COALESCE(entity_id, '00000000-0000-0000-0000-000000000000'::uuid)) DO NOTHING
+                """,
+                all_links
+            )
+            _log(log_buffer, f"      [8.3] Insert {len(all_links)} semantic links: {time_mod.time() - insert_start:.3f}s")
+
+    except Exception as e:
+        logger.error(f"Failed to create semantic links: {str(e)}")
+        import traceback
+        traceback.print_exc()
+        raise
+
+
+async def insert_entity_links_batch(conn, links: List[tuple]):
+    """
+    Insert all entity links in a single batch.
+
+    Args:
+        conn: Database connection
+        links: List of tuples (from_unit_id, to_unit_id, link_type, weight, entity_id)
+    """
+    if not links:
+        return
+
+    try:
+        await conn.executemany(
+            """
+            INSERT INTO memory_links (from_unit_id, to_unit_id, link_type, weight, entity_id)
+            VALUES ($1, $2, $3, $4, $5)
+            ON CONFLICT (from_unit_id, to_unit_id, link_type, COALESCE(entity_id, '00000000-0000-0000-0000-000000000000'::uuid)) DO NOTHING
+            """,
+            links
+        )
+    except Exception as e:
+        logger.warning(f"Failed to insert entity links: {str(e)}")
+
+
+async def create_causal_links_batch(
+    conn,
+    unit_ids: List[str],
+    causal_relations_per_fact: List[List[dict]],
+) -> int:
+    """
+    Create causal links between facts based on LLM-extracted causal relationships.
+
+    Args:
+        conn: Database connection
+        unit_ids: List of unit IDs (in same order as causal_relations_per_fact)
+        causal_relations_per_fact: List of causal relations for each fact.
+            Each element is a list of dicts with:
+            - target_fact_index: Index into unit_ids for the target fact
+            - relation_type: "causes", "caused_by", "enables", or "prevents"
+            - strength: Float in [0.0, 1.0] representing relationship strength
+
+    Returns:
+        Number of causal links created
+
+    Causal link types:
+    - "causes": This fact directly causes the target fact (forward causation)
+    - "caused_by": This fact was caused by the target fact (backward causation)
+    - "enables": This fact enables/allows the target fact (enablement)
+    - "prevents": This fact prevents/blocks the target fact (prevention)
+    """
+    if not unit_ids or not causal_relations_per_fact:
+        return 0
+
+    try:
+        import time as time_mod
+        create_start = time_mod.time()
+
+        # Build links list
+        links = []
+        for fact_idx, causal_relations in enumerate(causal_relations_per_fact):
+            if not causal_relations:
+                continue
+
+            from_unit_id = unit_ids[fact_idx]
+
+            for relation in causal_relations:
+                target_idx = relation['target_fact_index']
+                relation_type = relation['relation_type']
+                strength = relation.get('strength', 1.0)
+
+                # Validate target index
+                if target_idx < 0 or target_idx >= len(unit_ids):
+                    logger.warning(f"Invalid target_fact_index {target_idx} in causal relation from fact {fact_idx}")
+                    continue
+
+                to_unit_id = unit_ids[target_idx]
+
+                # Don't create self-links
+                if from_unit_id == to_unit_id:
+                    continue
+
+                # Add the causal link
+                # link_type is the relation_type (e.g., "causes", "caused_by")
+                # weight is the strength of the relationship
+                links.append((from_unit_id, to_unit_id, relation_type, strength, None))
+
+        logger.debug(f"Generated {len(links)} causal links in {time_mod.time() - create_start:.3f}s")
+
+        if links:
+            insert_start = time_mod.time()
+            await conn.executemany(
+                """
+                INSERT INTO memory_links (from_unit_id, to_unit_id, link_type, weight, entity_id)
+                VALUES ($1, $2, $3, $4, $5)
+                ON CONFLICT (from_unit_id, to_unit_id, link_type, COALESCE(entity_id, '00000000-0000-0000-0000-000000000000'::uuid)) DO NOTHING
+                """,
+                links
+            )
+            logger.debug(f"Inserted {len(links)} causal links in {time_mod.time() - insert_start:.3f}s")
+
+        return len(links)
+
+    except Exception as e:
+        logger.error(f"Failed to create causal links: {str(e)}")
+        import traceback
+        traceback.print_exc()
+        raise

hindsight-api/hindsight_api/engine/llm_wrapper.py

@@ -182,10 +182,10 @@ class LLMConfig:
     @classmethod
     def for_memory(cls) -> "LLMConfig":
         """Create configuration for memory operations from environment variables."""
-        provider = os.getenv("MEMORA_API_LLM_PROVIDER", "groq")
-        api_key = os.getenv("MEMORA_API_LLM_API_KEY")
-        base_url = os.getenv("MEMORA_API_LLM_BASE_URL")
-        model = os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-120b")
+        provider = os.getenv("HINDSIGHT_API_LLM_PROVIDER", "groq")
+        api_key = os.getenv("HINDSIGHT_API_LLM_API_KEY")
+        base_url = os.getenv("HINDSIGHT_API_LLM_BASE_URL")
+        model = os.getenv("HINDSIGHT_API_LLM_MODEL", "openai/gpt-oss-120b")
 
         # Set default base URL if not provided
         if not base_url:
@@ -211,10 +211,10 @@ class LLMConfig:
         Falls back to memory LLM config if judge-specific config not set.
         """
         # Check if judge-specific config exists, otherwise fall back to memory config
-        provider = os.getenv("MEMORA_API_JUDGE_LLM_PROVIDER", os.getenv("MEMORA_API_LLM_PROVIDER", "groq"))
-        api_key = os.getenv("MEMORA_API_JUDGE_LLM_API_KEY", os.getenv("MEMORA_API_LLM_API_KEY"))
-        base_url = os.getenv("MEMORA_API_JUDGE_LLM_BASE_URL", os.getenv("MEMORA_API_LLM_BASE_URL"))
-        model = os.getenv("MEMORA_API_JUDGE_LLM_MODEL", os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-120b"))
+        provider = os.getenv("HINDSIGHT_API_JUDGE_LLM_PROVIDER", os.getenv("HINDSIGHT_API_LLM_PROVIDER", "groq"))
+        api_key = os.getenv("HINDSIGHT_API_JUDGE_LLM_API_KEY", os.getenv("HINDSIGHT_API_LLM_API_KEY"))
+        base_url = os.getenv("HINDSIGHT_API_JUDGE_LLM_BASE_URL", os.getenv("HINDSIGHT_API_LLM_BASE_URL"))
+        model = os.getenv("HINDSIGHT_API_JUDGE_LLM_MODEL", os.getenv("HINDSIGHT_API_LLM_MODEL", "openai/gpt-oss-120b"))
 
         # Set default base URL if not provided
         if not base_url:

hindsight-api/hindsight_api/engine/memory_engine.py

@@ -1,5 +1,5 @@
 """
-Temporal + Semantic + Entity Memory System for AI Agents.
+Memory Engine for AI Agents.
 
 This implements a sophisticated memory architecture that combines:
 1. Temporal links: Memories connected by time proximity
@@ -8,6 +8,7 @@ This implements a sophisticated memory architecture that combines:
 4. Spreading activation: Search through the graph with activation decay
 5. Dynamic weighting: Recency and frequency-based importance
 """
+import json
 import os
 from datetime import datetime, timedelta, timezone
 from typing import Any, Dict, List, Optional, Tuple, Union
@@ -19,18 +20,26 @@ import time
 import numpy as np
 import uuid
 import logging
+from pydantic import BaseModel, Field
 
+from .query_analyzer import QueryAnalyzer
 from .utils import (
     extract_facts,
     calculate_recency_weight,
     calculate_frequency_weight,
 )
 from .entity_resolver import EntityResolver
-from .operations import EmbeddingOperationsMixin, LinkOperationsMixin, ThinkOperationsMixin, AgentOperationsMixin
+from . import (
+    embedding_utils,
+    link_utils,
+    think_utils,
+    agent_utils,
+)
 from .llm_wrapper import LLMConfig
 from .response_models import SearchResult as SearchResultModel, ThinkResult, MemoryFact
 from .task_backend import TaskBackend, AsyncIOQueueBackend
 from .search.reranking import CrossEncoderReranker
+from ..pg0 import EmbeddedPostgres
 
 
 def utcnow():
@@ -41,8 +50,10 @@ def utcnow():
 # Logger for memory system
 logger = logging.getLogger(__name__)
 
-# Tiktoken for token budget filtering
+from .db_utils import acquire_with_retry, retry_with_backoff
+
 import tiktoken
+from dateutil import parser as date_parser
 
 # Cache tiktoken encoding for token budget filtering (module-level singleton)
 _TIKTOKEN_ENCODING = None
@@ -55,20 +66,15 @@ def _get_tiktoken_encoding():
     return _TIKTOKEN_ENCODING
 
 
-class TemporalSemanticMemory(
-    EmbeddingOperationsMixin,
-    LinkOperationsMixin,
-    ThinkOperationsMixin,
-    AgentOperationsMixin,
-):
+class MemoryEngine:
     """
     Advanced memory system using temporal and semantic linking with PostgreSQL.
 
-    Uses mixin architecture for code organization:
-    - EmbeddingOperationsMixin: Embedding generation
-    - LinkOperationsMixin: Entity, temporal, and semantic link creation
-    - ThinkOperationsMixin: Think operations for formulating answers with opinions
-    - AgentOperationsMixin: Agent profile and personality management
+    This class provides:
+    - Embedding generation for semantic search
+    - Entity, temporal, and semantic link creation
+    - Think operations for formulating answers with opinions
+    - Agent profile and personality management
     """
 
     def __init__(
@@ -80,7 +86,7 @@ class TemporalSemanticMemory(
         memory_llm_base_url: Optional[str] = None,
         embeddings: Optional[Embeddings] = None,
         cross_encoder: Optional[CrossEncoderModel] = None,
-        query_analyzer: Optional["QueryAnalyzer"] = None,
+        query_analyzer: Optional[QueryAnalyzer] = None,
         pool_min_size: int = 5,
         pool_max_size: int = 100,
         task_backend: Optional[TaskBackend] = None,
@@ -104,8 +110,15 @@ class TemporalSemanticMemory(
                           Increase for parallel think/search operations (e.g., 200-300 for 100+ parallel thinks)
             task_backend: Custom task backend for async task execution. If not provided, uses AsyncIOQueueBackend
         """
+        # Track pg0 instance (if used)
+        self._pg0: Optional[EmbeddedPostgres] = None
+
         # Initialize PostgreSQL connection URL
-        self.db_url = db_url
+        # "pg0" or "embedded-pg" are special values that trigger embedded PostgreSQL via pg0
+        # The actual URL will be set during initialize() after starting the server
+        self._use_pg0 = db_url in ("pg0", "embedded-pg")
+        self.db_url = db_url if not self._use_pg0 else None
+
 
         # Set default base URL if not provided
         if memory_llm_base_url is None:
@@ -135,7 +148,7 @@ class TemporalSemanticMemory(
         if query_analyzer is not None:
             self.query_analyzer = query_analyzer
         else:
-            from memora.query_analyzer import TransformerQueryAnalyzer
+            from .query_analyzer import TransformerQueryAnalyzer
             self.query_analyzer = TransformerQueryAnalyzer()
 
         # Initialize LLM configuration
@@ -188,7 +201,7 @@ class TemporalSemanticMemory(
         try:
             # Convert string UUIDs to UUID type for faster matching
             uuid_list = [uuid.UUID(nid) for nid in node_ids]
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 await conn.execute(
                     "UPDATE memory_units SET access_count = access_count + 1 WHERE id = ANY($1::uuid[])",
                     uuid_list
@@ -242,7 +255,7 @@ class TemporalSemanticMemory(
         if operation_id:
             try:
                 pool = await self._get_pool()
-                async with pool.acquire() as conn:
+                async with acquire_with_retry(pool) as conn:
                     result = await conn.fetchrow(
                         "SELECT id FROM async_operations WHERE id = $1",
                         uuid.UUID(operation_id)
@@ -297,7 +310,7 @@ class TemporalSemanticMemory(
         """Helper to delete an operation record from the database."""
         try:
             pool = await self._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 await conn.execute(
                     "DELETE FROM async_operations WHERE id = $1",
                     uuid.UUID(operation_id)
@@ -314,7 +327,7 @@ class TemporalSemanticMemory(
             full_error = f"{error_message}\n\nTraceback:\n{error_traceback}"
             truncated_error = full_error[:5000] if len(full_error) > 5000 else full_error
 
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 await conn.execute(
                     """
                     UPDATE async_operations
@@ -333,6 +346,13 @@ class TemporalSemanticMemory(
         if self._initialized:
             return
 
+        # Start pg0 embedded PostgreSQL if configured
+        if self._use_pg0:
+            logger.info("Starting pg0 embedded PostgreSQL...")
+            self._pg0 = EmbeddedPostgres()
+            self.db_url = await self._pg0.ensure_running()
+            logger.info(f"pg0 PostgreSQL running at: {self.db_url}")
+
         # Create connection pool
         # For read-heavy workloads with many parallel think/search operations,
         # we need a larger pool. Read operations don't need strong isolation.
@@ -341,7 +361,8 @@ class TemporalSemanticMemory(
             min_size=self._pool_min_size,
             max_size=self._pool_max_size,
             command_timeout=60,
-            statement_cache_size=0  # Disable prepared statement cache
+            statement_cache_size=0,  # Disable prepared statement cache
+            timeout=30,  # Connection acquisition timeout (seconds)
         )
 
         # Initialize entity resolver with pool
@@ -360,6 +381,20 @@ class TemporalSemanticMemory(
             await self.initialize()
         return self._pool
 
+    async def _acquire_connection(self):
+        """
+        Acquire a connection from the pool with retry logic.
+
+        Returns an async context manager that yields a connection.
+        Retries on transient connection errors with exponential backoff.
+        """
+        pool = await self._get_pool()
+
+        async def acquire():
+            return await pool.acquire()
+
+        return await _retry_with_backoff(acquire)
+
     async def close(self):
         """Close the connection pool and shutdown background workers."""
         logger.info("close() started")
@@ -379,6 +414,14 @@ class TemporalSemanticMemory(
             logger.debug("no pool to close")
 
         self._initialized = False
+
+        # Stop pg0 if we started it
+        if self._pg0 is not None:
+            logger.info("Stopping pg0...")
+            await self._pg0.stop()
+            self._pg0 = None
+            logger.info("pg0 stopped")
+
         logger.debug("close() completed")
 
     async def wait_for_background_tasks(self):
@@ -729,7 +772,8 @@ class TemporalSemanticMemory(
             log_buffer.append(f"{'='*60}")
 
             # Get agent name for fact extraction
-            profile = await self.get_agent_profile(agent_id)
+            pool = await self._get_pool()
+            profile = await agent_utils.get_agent_profile(pool, agent_id)
             agent_name = profile["name"]
 
             # Step 1: Extract facts from ALL contents in parallel
@@ -774,7 +818,7 @@ class TemporalSemanticMemory(
                     all_fact_texts.append(fact_dict['fact'])
 
                     # Extract temporal fields (new schema with ranges)
-                    from dateutil import parser as date_parser
+
                     try:
                         # Try new schema first (occurred_start/end)
                         occurred_start = date_parser.isoparse(fact_dict['occurred_start'])
@@ -855,14 +899,14 @@ class TemporalSemanticMemory(
 
             # Step 2b: Generate ALL embeddings in ONE batch using augmented texts (HUGE speedup!)
             step_start = time.time()
-            all_embeddings = await self._generate_embeddings_batch(augmented_texts)
+            all_embeddings = await embedding_utils.generate_embeddings_batch(self.embeddings, augmented_texts)
             log_buffer.append(f"[2] Generate embeddings (parallel): {len(all_embeddings)} embeddings in {time.time() - step_start:.3f}s")
 
             # Step 3: Process everything in ONE database transaction
             logger.debug("Getting connection pool")
             pool = await self._get_pool()
             logger.debug("Acquiring connection from pool")
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 logger.debug("Starting transaction")
                 async with conn.transaction():
                     logger.debug("Inside transaction")
@@ -1035,8 +1079,8 @@ class TemporalSemanticMemory(
                         # Process entities for ALL units
                         logger.debug("Processing entities")
                         step_start = time.time()
-                        all_entity_links = await self._extract_entities_batch_optimized(
-                            conn, agent_id, created_unit_ids, filtered_sentences, "", filtered_dates, filtered_entities, log_buffer
+                        all_entity_links = await link_utils.extract_entities_batch_optimized(
+                            self.entity_resolver, conn, agent_id, created_unit_ids, filtered_sentences, "", filtered_dates, filtered_entities, log_buffer
                         )
                         logger.debug(f"Entity processing complete: {len(all_entity_links)} links")
                         log_buffer.append(f"[6] Process entities (batched): {time.time() - step_start:.3f}s")
@@ -1044,14 +1088,14 @@ class TemporalSemanticMemory(
                         # Create temporal links
                         logger.debug("Creating temporal links")
                         step_start = time.time()
-                        await self._create_temporal_links_batch_per_fact(conn, agent_id, created_unit_ids, log_buffer=log_buffer)
+                        await link_utils.create_temporal_links_batch_per_fact(conn, agent_id, created_unit_ids, log_buffer=log_buffer)
                         logger.debug("Temporal links complete")
                         log_buffer.append(f"[7] Batch create temporal links: {time.time() - step_start:.3f}s")
 
                         # Create semantic links
                         logger.debug("Creating semantic links")
                         step_start = time.time()
-                        await self._create_semantic_links_batch(conn, agent_id, created_unit_ids, filtered_embeddings, log_buffer=log_buffer)
+                        await link_utils.create_semantic_links_batch(conn, agent_id, created_unit_ids, filtered_embeddings, log_buffer=log_buffer)
                         logger.debug("Semantic links complete")
                         log_buffer.append(f"[8] Batch create semantic links: {time.time() - step_start:.3f}s")
 
@@ -1059,14 +1103,14 @@ class TemporalSemanticMemory(
                         logger.debug("Inserting entity links")
                         step_start = time.time()
                         if all_entity_links:
-                            await self._insert_entity_links_batch(conn, all_entity_links)
+                            await link_utils.insert_entity_links_batch(conn, all_entity_links)
                         logger.debug("Entity links inserted")
                         log_buffer.append(f"[9] Batch insert entity links: {time.time() - step_start:.3f}s")
 
                         # Create causal links
                         logger.debug("Creating causal links")
                         step_start = time.time()
-                        causal_link_count = await self._create_causal_links_batch(
+                        causal_link_count = await link_utils.create_causal_links_batch(
                             conn, created_unit_ids, filtered_causal_relations
                         )
                         logger.debug(f"Causal links complete: {causal_link_count} links created")
@@ -1261,7 +1305,7 @@ class TemporalSemanticMemory(
         try:
             # Step 1: Generate query embedding (for semantic search)
             step_start = time.time()
-            query_embedding = self._generate_embedding(query)
+            query_embedding = embedding_utils.generate_embedding(self.embeddings, query)
             step_duration = time.time() - step_start
             log_buffer.append(f"  [1] Generate query embedding: {step_duration:.3f}s")
 
@@ -1593,7 +1637,7 @@ class TemporalSemanticMemory(
             Dictionary with document info or None if not found
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             doc = await conn.fetchrow(
                 """
                 SELECT d.id, d.agent_id, d.original_text, d.content_hash,
@@ -1631,7 +1675,7 @@ class TemporalSemanticMemory(
             Dictionary with counts of deleted items
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             async with conn.transaction():
                 # Count units before deletion
                 units_count = await conn.fetchval(
@@ -1666,7 +1710,7 @@ class TemporalSemanticMemory(
             Dictionary with deletion result
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             async with conn.transaction():
                 # Delete the memory unit (cascades to links and associations)
                 deleted = await conn.fetchval(
@@ -1700,7 +1744,7 @@ class TemporalSemanticMemory(
             Dictionary with counts of deleted items
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             async with conn.transaction():
                 try:
                     if fact_type:
@@ -1751,7 +1795,7 @@ class TemporalSemanticMemory(
             Dict with nodes, edges, and table_rows
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             # Get memory units, optionally filtered by agent_id and fact_type
             query_conditions = []
             query_params = []
@@ -1922,7 +1966,7 @@ class TemporalSemanticMemory(
             Dict with items (list of memory units) and total count
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             # Build query conditions
             query_conditions = []
             query_params = []
@@ -2036,7 +2080,7 @@ class TemporalSemanticMemory(
             Dict with items (list of documents without original_text) and total count
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             # Build query conditions
             query_conditions = []
             query_params = []
@@ -2153,7 +2197,7 @@ class TemporalSemanticMemory(
             Dict with document details including original_text, or None if not found
         """
         pool = await self._get_pool()
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             doc = await conn.fetchrow("""
                 SELECT
                     id,
@@ -2336,7 +2380,7 @@ Guidelines:
             logger.debug(f"[REINFORCE] Starting opinion reinforcement for {len(entity_names)} entities")
 
             pool = await self._get_pool()
-            async with pool.acquire() as conn:
+            async with acquire_with_retry(pool) as conn:
                 # Find all opinions related to these entities
                 opinions = await conn.fetch(
                     """
@@ -2439,3 +2483,229 @@ Guidelines:
             import traceback
             traceback.print_exc()
 
+    # ==================== Agent Profile Methods ====================
+
+    async def get_agent_profile(self, agent_id: str) -> Dict:
+        """
+        Get agent profile (name, personality + background).
+        Auto-creates agent with default values if not exists.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            Dict with 'name' (str), 'personality' (dict) and 'background' (str) keys
+        """
+        pool = await self._get_pool()
+        return await agent_utils.get_agent_profile(pool, agent_id)
+
+    async def update_agent_personality(
+        self,
+        agent_id: str,
+        personality: Dict[str, float]
+    ) -> None:
+        """
+        Update agent personality traits.
+
+        Args:
+            agent_id: Agent identifier
+            personality: Dict with Big Five traits + bias_strength (all 0-1)
+        """
+        pool = await self._get_pool()
+        await agent_utils.update_agent_personality(pool, agent_id, personality)
+
+    async def merge_agent_background(
+        self,
+        agent_id: str,
+        new_info: str,
+        update_personality: bool = True
+    ) -> dict:
+        """
+        Merge new background information with existing background using LLM.
+        Normalizes to first person ("I") and resolves conflicts.
+        Optionally infers personality traits from the merged background.
+
+        Args:
+            agent_id: Agent identifier
+            new_info: New background information to add/merge
+            update_personality: If True, infer Big Five traits from background (default: True)
+
+        Returns:
+            Dict with 'background' (str) and optionally 'personality' (dict) keys
+        """
+        pool = await self._get_pool()
+        return await agent_utils.merge_agent_background(
+            pool, self._llm_config, agent_id, new_info, update_personality
+        )
+
+    async def list_agents(self) -> list:
+        """
+        List all agents in the system.
+
+        Returns:
+            List of dicts with agent_id, name, personality, background, created_at, updated_at
+        """
+        pool = await self._get_pool()
+        return await agent_utils.list_agents(pool)
+
+    # ==================== Think Methods ====================
+
+    async def think_async(
+        self,
+        agent_id: str,
+        query: str,
+        thinking_budget: int = 50,
+        context: str = None,
+    ) -> ThinkResult:
+        """
+        Think and formulate an answer using agent identity, world facts, and opinions.
+
+        This method:
+        1. Retrieves agent facts (agent's identity and past actions)
+        2. Retrieves world facts (general knowledge)
+        3. Retrieves existing opinions (agent's formed perspectives)
+        4. Uses LLM to formulate an answer
+        5. Extracts and stores any new opinions formed during thinking
+        6. Returns plain text answer and the facts used
+
+        Args:
+            agent_id: Agent identifier
+            query: Question to answer
+            thinking_budget: Number of memory units to explore
+            context: Additional context string to include in LLM prompt (not used in search)
+
+        Returns:
+            ThinkResult containing:
+                - text: Plain text answer (no markdown)
+                - based_on: Dict with 'world', 'agent', and 'opinion' fact lists (MemoryFact objects)
+                - new_opinions: List of newly formed opinions
+        """
+        # Use cached LLM config
+        if self._llm_config is None:
+            raise ValueError("Memory LLM API key not set. Set HINDSIGHT_API_LLM_API_KEY environment variable.")
+
+        # Steps 1-3: Run multi-fact-type search (12-way retrieval: 4 methods × 3 fact types)
+        search_result = await self.search_async(
+            agent_id=agent_id,
+            query=query,
+            thinking_budget=thinking_budget,
+            max_tokens=4096,
+            enable_trace=False,
+            fact_type=['agent', 'world', 'opinion']
+        )
+
+        all_results = search_result.results
+        logger.info(f"[THINK] Search returned {len(all_results)} results")
+
+        # Split results by fact type for structured response
+        agent_results = [r for r in all_results if r.fact_type == 'agent']
+        world_results = [r for r in all_results if r.fact_type == 'world']
+        opinion_results = [r for r in all_results if r.fact_type == 'opinion']
+
+        logger.info(f"[THINK] Split results - agent: {len(agent_results)}, world: {len(world_results)}, opinion: {len(opinion_results)}")
+
+        # Format facts for LLM
+        agent_facts_text = think_utils.format_facts_for_prompt(agent_results)
+        world_facts_text = think_utils.format_facts_for_prompt(world_results)
+        opinion_facts_text = think_utils.format_facts_for_prompt(opinion_results)
+
+        logger.info(f"[THINK] Formatted facts - agent: {len(agent_facts_text)} chars, world: {len(world_facts_text)} chars, opinion: {len(opinion_facts_text)} chars")
+
+        # Get agent profile (name, personality + background)
+        profile = await self.get_agent_profile(agent_id)
+        name = profile["name"]
+        personality = profile["personality"]
+        background = profile["background"]
+
+        # Build the prompt
+        prompt = think_utils.build_think_prompt(
+            agent_facts_text=agent_facts_text,
+            world_facts_text=world_facts_text,
+            opinion_facts_text=opinion_facts_text,
+            query=query,
+            name=name,
+            personality=personality,
+            background=background,
+            context=context,
+        )
+
+        logger.info(f"[THINK] Full prompt length: {len(prompt)} chars")
+        logger.debug(f"[THINK] Prompt preview (first 500 chars): {prompt[:500]}")
+
+        system_message = think_utils.get_system_message(personality)
+
+        answer_text = await self._llm_config.call(
+            messages=[
+                {"role": "system", "content": system_message},
+                {"role": "user", "content": prompt}
+            ],
+            scope="memory_think",
+            temperature=0.9,
+            max_tokens=1000
+        )
+
+        answer_text = answer_text.strip()
+
+        # Submit form_opinion task for background processing
+        logger.debug(f"[THINK] Submitting form_opinion task for agent {agent_id}")
+        await self._task_backend.submit_task({
+            'type': 'form_opinion',
+            'agent_id': agent_id,
+            'answer_text': answer_text,
+            'query': query
+        })
+        logger.debug(f"[THINK] form_opinion task submitted")
+
+        # Return response with facts split by type
+        return ThinkResult(
+            text=answer_text,
+            based_on={
+                "world": world_results,
+                "agent": agent_results,
+                "opinion": opinion_results
+            },
+            new_opinions=[]  # Opinions are being extracted asynchronously
+        )
+
+    async def _extract_and_store_opinions_async(
+        self,
+        agent_id: str,
+        answer_text: str,
+        query: str
+    ):
+        """
+        Background task to extract and store opinions from think response.
+
+        This runs asynchronously and does not block the think response.
+
+        Args:
+            agent_id: Agent identifier
+            answer_text: The generated answer text
+            query: The original query
+        """
+        try:
+            logger.debug(f"[THINK] Extracting opinions from answer for agent {agent_id}")
+            # Extract opinions from the answer
+            new_opinions = await think_utils.extract_opinions_from_text(
+                self._llm_config, text=answer_text, query=query
+            )
+            logger.debug(f"[THINK] Extracted {len(new_opinions)} opinions")
+
+            # Store new opinions
+            if new_opinions:
+                from datetime import datetime, timezone
+                current_time = datetime.now(timezone.utc)
+                for opinion_dict in new_opinions:
+                    await self.put_async(
+                        agent_id=agent_id,
+                        content=opinion_dict["text"],
+                        context=f"formed during thinking about: {query}",
+                        event_date=current_time,
+                        fact_type_override='opinion',
+                        confidence_score=opinion_dict["confidence"]
+                    )
+
+                logger.debug(f"[THINK] Extracted and stored {len(new_opinions)} new opinions")
+        except Exception as e:
+            logger.warning(f"[THINK] Failed to extract/store opinions: {str(e)}")
+

hindsight-api/hindsight_api/engine/response_models.py

@@ -1,13 +1,13 @@
 """
-Core response models for Memora memory system.
+Core response models for Hindsight memory system.
 
-These models define the structure of data returned by the core TemporalSemanticMemory class.
+These models define the structure of data returned by the core MemoryEngine class.
 API response models should be kept separate and convert from these core models to maintain
 API stability even if internal models change.
 """
 
 from typing import Optional, List, Dict, Any
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, ConfigDict
 
 
 class MemoryFact(BaseModel):
@@ -17,6 +17,19 @@ class MemoryFact(BaseModel):
     This represents a unit of information stored in the memory system,
     including both the content and metadata.
     """
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "id": "123e4567-e89b-12d3-a456-426614174000",
+            "text": "Alice works at Google on the AI team",
+            "fact_type": "world",
+            "context": "work info",
+            "event_date": "2024-01-15T10:30:00Z",
+            "document_id": "session_abc123",
+            "metadata": {"source": "slack"},
+            "activation": 0.95
+        }
+    })
+
     id: str = Field(description="Unique identifier for the memory fact")
     text: str = Field(description="The actual text content of the memory")
     fact_type: str = Field(description="Type of fact: 'world', 'agent', or 'opinion'")
@@ -31,20 +44,6 @@ class MemoryFact(BaseModel):
     # Internal metrics (used by system but may not be exposed in API)
     activation: Optional[float] = Field(None, description="Internal activation score")
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "id": "123e4567-e89b-12d3-a456-426614174000",
-                "text": "Alice works at Google on the AI team",
-                "fact_type": "world",
-                "context": "work info",
-                "event_date": "2024-01-15T10:30:00Z",
-                "document_id": "session_abc123",
-                "metadata": {"source": "slack"},
-                "activation": 0.95
-            }
-        }
-
 
 class SearchResult(BaseModel):
     """
@@ -53,28 +52,27 @@ class SearchResult(BaseModel):
     Contains a list of matching memory facts and optional trace information
     for debugging and transparency.
     """
-    results: List[MemoryFact] = Field(description="List of memory facts matching the query")
-    trace: Optional[Dict[str, Any]] = Field(None, description="Trace information for debugging")
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "results": [
-                    {
-                        "id": "123e4567-e89b-12d3-a456-426614174000",
-                        "text": "Alice works at Google on the AI team",
-                        "fact_type": "world",
-                        "context": "work info",
-                        "event_date": "2024-01-15T10:30:00Z",
-                        "activation": 0.95
-                    }
-                ],
-                "trace": {
-                    "query": "What did Alice say about machine learning?",
-                    "num_results": 1
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "results": [
+                {
+                    "id": "123e4567-e89b-12d3-a456-426614174000",
+                    "text": "Alice works at Google on the AI team",
+                    "fact_type": "world",
+                    "context": "work info",
+                    "event_date": "2024-01-15T10:30:00Z",
+                    "activation": 0.95
                 }
+            ],
+            "trace": {
+                "query": "What did Alice say about machine learning?",
+                "num_results": 1
             }
         }
+    })
+
+    results: List[MemoryFact] = Field(description="List of memory facts matching the query")
+    trace: Optional[Dict[str, Any]] = Field(None, description="Trace information for debugging")
 
 
 class ThinkResult(BaseModel):
@@ -84,6 +82,28 @@ class ThinkResult(BaseModel):
     Contains the formulated answer, the facts it was based on (organized by type),
     and any new opinions that were formed during the thinking process.
     """
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "text": "Based on my knowledge, machine learning is being actively used in healthcare...",
+            "based_on": {
+                "world": [
+                    {
+                        "id": "123e4567-e89b-12d3-a456-426614174000",
+                        "text": "Machine learning is used in medical diagnosis",
+                        "fact_type": "world",
+                        "context": "healthcare",
+                        "event_date": "2024-01-15T10:30:00Z"
+                    }
+                ],
+                "agent": [],
+                "opinion": []
+            },
+            "new_opinions": [
+                "Machine learning has great potential in healthcare"
+            ]
+        }
+    })
+
     text: str = Field(description="The formulated answer text")
     based_on: Dict[str, List[MemoryFact]] = Field(
         description="Facts used to formulate the answer, organized by type (world, agent, opinion)"
@@ -93,29 +113,6 @@ class ThinkResult(BaseModel):
         description="List of newly formed opinions during thinking"
     )
 
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "text": "Based on my knowledge, machine learning is being actively used in healthcare...",
-                "based_on": {
-                    "world": [
-                        {
-                            "id": "123e4567-e89b-12d3-a456-426614174000",
-                            "text": "Machine learning is used in medical diagnosis",
-                            "fact_type": "world",
-                            "context": "healthcare",
-                            "event_date": "2024-01-15T10:30:00Z"
-                        }
-                    ],
-                    "agent": [],
-                    "opinion": []
-                },
-                "new_opinions": [
-                    "Machine learning has great potential in healthcare"
-                ]
-            }
-        }
-
 
 class Opinion(BaseModel):
     """
@@ -124,13 +121,12 @@ class Opinion(BaseModel):
     Opinions represent the agent's formed perspectives on topics,
     with a confidence level indicating strength of belief.
     """
+    model_config = ConfigDict(json_schema_extra={
+        "example": {
+            "text": "Machine learning has great potential in healthcare",
+            "confidence": 0.85
+        }
+    })
+
     text: str = Field(description="The opinion text")
     confidence: float = Field(description="Confidence score between 0.0 and 1.0")
-
-    class Config:
-        json_schema_extra = {
-            "example": {
-                "text": "Machine learning has great potential in healthcare",
-                "confidence": 0.85
-            }
-        }

hindsight-api/hindsight_api/engine/search/reranking.py

@@ -24,7 +24,7 @@ class CrossEncoderReranker:
                           SentenceTransformersCrossEncoder with ms-marco-MiniLM-L-6-v2
         """
         if cross_encoder is None:
-            from ..cross_encoder import SentenceTransformersCrossEncoder
+            from hindsight_api.engine.cross_encoder import SentenceTransformersCrossEncoder
             cross_encoder = SentenceTransformersCrossEncoder()
         self.cross_encoder = cross_encoder
 

hindsight-api/hindsight_api/engine/search/retrieval.py

@@ -11,6 +11,7 @@ Implements:
 from typing import List, Dict, Any, Tuple, Optional
 from datetime import datetime
 import asyncio
+from ..db_utils import acquire_with_retry
 
 
 async def retrieve_semantic(
@@ -241,11 +242,6 @@ async def retrieve_temporal(
     if end_date.tzinfo is None:
         end_date = end_date.replace(tzinfo=timezone.utc)
 
-    # Find entry points: facts in date range with semantic relevance
-    import logging
-    logger = logging.getLogger(__name__)
-    logger.info(f"Temporal retrieval: searching for facts between {start_date} and {end_date} (agent={agent_id}, fact_type={fact_type})")
-
     entry_points = await conn.fetch(
         """
         SELECT id, text, context, event_date, occurred_start, occurred_end, mentioned_at, access_count, embedding, fact_type, document_id,
@@ -274,8 +270,6 @@ async def retrieve_temporal(
         query_emb_str, agent_id, fact_type, start_date, end_date, semantic_threshold
     )
 
-    logger.info(f"Temporal retrieval: found {len(entry_points)} entry points")
-
     if not entry_points:
         # Check if there are ANY memories with temporal metadata for this agent
         total_with_dates = await conn.fetchval(
@@ -284,7 +278,6 @@ async def retrieve_temporal(
                AND (occurred_start IS NOT NULL OR occurred_end IS NOT NULL OR mentioned_at IS NOT NULL)""",
             agent_id, fact_type
         )
-        logger.info(f"Temporal retrieval: agent has {total_with_dates} total memories with temporal metadata (fact_type={fact_type})")
         return []
 
     # Calculate temporal scores for entry points
@@ -442,26 +435,20 @@ async def retrieve_parallel(
         query_text, reference_date=question_date, analyzer=query_analyzer
     )
 
-    if temporal_constraint:
-        logger.info(f"Temporal constraint detected in retrieve_parallel: {temporal_constraint[0]} to {temporal_constraint[1]}")
-    else:
-        logger.info("No temporal constraint in retrieve_parallel")
-
-    # Each retrieval needs its own connection
     async def run_semantic():
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             return await retrieve_semantic(conn, query_embedding_str, agent_id, fact_type, limit=thinking_budget)
 
     async def run_bm25():
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             return await retrieve_bm25(conn, query_text, agent_id, fact_type, limit=thinking_budget)
 
     async def run_graph():
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             return await retrieve_graph(conn, query_embedding_str, agent_id, fact_type, budget=thinking_budget)
 
     async def run_temporal(start_date, end_date):
-        async with pool.acquire() as conn:
+        async with acquire_with_retry(pool) as conn:
             return await retrieve_temporal(
                 conn, query_embedding_str, agent_id, fact_type,
                 start_date, end_date, budget=thinking_budget, semantic_threshold=0.4

hindsight-api/hindsight_api/engine/search/temporal_extraction.py

@@ -7,7 +7,7 @@ Handles natural language temporal expressions using transformer-based query anal
 from typing import Optional, Tuple
 from datetime import datetime
 import logging
-from memora.query_analyzer import QueryAnalyzer, TransformerQueryAnalyzer
+from hindsight_api.engine.query_analyzer import QueryAnalyzer, TransformerQueryAnalyzer
 
 logger = logging.getLogger(__name__)
 
@@ -59,8 +59,6 @@ def extract_temporal_constraint(
             analysis.temporal_constraint.start_date,
             analysis.temporal_constraint.end_date
         )
-        logger.info(f"Temporal constraint extracted: {result[0].strftime('%Y-%m-%d')} to {result[1].strftime('%Y-%m-%d')}")
         return result
 
-    logger.info("No temporal constraint found in query")
     return None

hindsight-api/hindsight_api/engine/task_backend.py

@@ -23,7 +23,7 @@ class TaskBackend(ABC):
     2. Execute tasks through a provided executor callback
 
     The backend treats tasks as pure dictionaries that can be serialized
-    and sent over the network. The executor (typically TemporalSemanticMemory.execute_task)
+    and sent over the network. The executor (typically MemoryEngine.execute_task)
     receives the dict and routes it to the appropriate handler.
     """
 

hindsight-api/hindsight_api/engine/think_utils.py

@@ -0,0 +1,255 @@
+"""
+Think operation utilities for formulating answers based on agent and world facts.
+"""
+
+import asyncio
+import logging
+import re
+from datetime import datetime, timezone
+from typing import Dict, List, Any
+from pydantic import BaseModel, Field
+
+from .response_models import ThinkResult, MemoryFact
+
+logger = logging.getLogger(__name__)
+
+
+class Opinion(BaseModel):
+    """An opinion formed by the agent."""
+    opinion: str = Field(description="The opinion or perspective with reasoning included")
+    confidence: float = Field(description="Confidence score for this opinion (0.0 to 1.0, where 1.0 is very confident)")
+
+
+class OpinionExtractionResponse(BaseModel):
+    """Response containing extracted opinions."""
+    opinions: List[Opinion] = Field(
+        default_factory=list,
+        description="List of opinions formed with their supporting reasons and confidence scores"
+    )
+
+
+def describe_trait(name: str, value: float) -> str:
+    """Convert trait value to descriptive text."""
+    if value >= 0.8:
+        return f"very high {name}"
+    elif value >= 0.6:
+        return f"high {name}"
+    elif value >= 0.4:
+        return f"moderate {name}"
+    elif value >= 0.2:
+        return f"low {name}"
+    else:
+        return f"very low {name}"
+
+
+def build_personality_description(personality: Dict) -> str:
+    """Build a personality description string from personality traits."""
+    return f"""Your personality traits:
+- {describe_trait('openness to new ideas', personality['openness'])}
+- {describe_trait('conscientiousness and organization', personality['conscientiousness'])}
+- {describe_trait('extraversion and sociability', personality['extraversion'])}
+- {describe_trait('agreeableness and cooperation', personality['agreeableness'])}
+- {describe_trait('emotional sensitivity', personality['neuroticism'])}
+
+Personality influence strength: {int(personality['bias_strength'] * 100)}% (how much your personality shapes your opinions)"""
+
+
+def format_facts_for_prompt(facts: List[MemoryFact]) -> str:
+    """Format facts as JSON for LLM prompt."""
+    import json
+
+    if not facts:
+        return "[]"
+    formatted = []
+    for fact in facts:
+        fact_obj = {
+            "text": fact.text
+        }
+
+        # Add context if available
+        if fact.context:
+            fact_obj["context"] = fact.context
+
+        # Add event_date if available
+        if fact.event_date:
+            event_date = fact.event_date
+            if isinstance(event_date, str):
+                fact_obj["event_date"] = event_date
+            elif isinstance(event_date, datetime):
+                fact_obj["event_date"] = event_date.strftime('%Y-%m-%d %H:%M:%S')
+
+        # Add activation if available
+        if fact.activation is not None:
+            fact_obj["score"] = fact.activation
+
+        formatted.append(fact_obj)
+
+    return json.dumps(formatted, indent=2)
+
+
+def build_think_prompt(
+    agent_facts_text: str,
+    world_facts_text: str,
+    opinion_facts_text: str,
+    query: str,
+    name: str,
+    personality: Dict,
+    background: str,
+    context: str = None,
+) -> str:
+    """Build the think prompt for the LLM."""
+    personality_desc = build_personality_description(personality)
+
+    name_section = f"""
+
+Your name: {name}
+"""
+
+    background_section = ""
+    if background:
+        background_section = f"""
+
+Your background:
+{background}
+"""
+
+    context_section = ""
+    if context:
+        context_section = f"""
+ADDITIONAL CONTEXT:
+{context}
+
+"""
+
+    return f"""Here's what I know and have experienced:
+
+MY IDENTITY & EXPERIENCES:
+{agent_facts_text}
+
+WHAT I KNOW ABOUT THE WORLD:
+{world_facts_text}
+
+MY EXISTING OPINIONS & BELIEFS:
+{opinion_facts_text}
+
+{context_section}{name_section}{personality_desc}{background_section}
+
+QUESTION: {query}
+
+Based on everything I know, believe, and who I am (including my name, personality and background), here's what I genuinely think about this question. I'll draw on my experiences, knowledge, opinions, and personal traits to give you my honest perspective."""
+
+
+def get_system_message(personality: Dict) -> str:
+    """Get the system message for the think LLM call."""
+    bias_strength = personality['bias_strength']
+    if bias_strength >= 0.7:
+        personality_instruction = "Your personality strongly influences your thinking. Let your traits guide how you interpret facts and form opinions. Don't be afraid to be biased based on your personality."
+    elif bias_strength >= 0.4:
+        personality_instruction = "Your personality moderately influences your thinking. Balance your personal traits with objective analysis."
+    else:
+        personality_instruction = "Your personality has minimal influence on your thinking. Focus primarily on facts while keeping your traits in mind."
+
+    return f"You are a person with your own thoughts, experiences, opinions, and personality. When asked a question, respond naturally like a human would - share what you genuinely think based on what you know and have experienced. {personality_instruction} Be direct, express your views confidently, and use 'I think', 'I believe', 'in my view', etc. Respond in plain text without markdown formatting."
+
+
+async def extract_opinions_from_text(
+    llm_config,
+    text: str,
+    query: str
+) -> List[Dict[str, Any]]:
+    """
+    Extract opinions with reasons and confidence from text using LLM.
+
+    Args:
+        llm_config: LLM configuration to use
+        text: Text to extract opinions from
+        query: The original query that prompted this response
+
+    Returns:
+        List of dicts with keys: 'text' (opinion with reasons), 'confidence' (score 0-1)
+    """
+    extraction_prompt = f"""Extract any NEW opinions or perspectives from the answer below and rewrite them in FIRST-PERSON as if YOU are stating the opinion directly.
+
+ORIGINAL QUESTION:
+{query}
+
+ANSWER PROVIDED:
+{text}
+
+Your task: Find opinions in the answer and rewrite them AS IF YOU ARE THE ONE SAYING THEM.
+
+An opinion is a judgment, viewpoint, or conclusion that goes beyond just stating facts.
+
+IMPORTANT: Do NOT extract statements like:
+- "I don't have enough information"
+- "The facts don't contain information about X"
+- "I cannot answer because..."
+
+ONLY extract actual opinions about substantive topics.
+
+CRITICAL FORMAT REQUIREMENTS:
+1. **ALWAYS start with first-person phrases**: "I think...", "I believe...", "In my view...", "I've come to believe...", "Previously I thought... but now..."
+2. **NEVER use third-person**: Do NOT say "The speaker thinks..." or "They believe..." - always use "I"
+3. Include the reasoning naturally within the statement
+4. Provide a confidence score (0.0 to 1.0)
+
+CORRECT Examples (✓ FIRST-PERSON):
+- "I think Alice is more reliable because she consistently delivers on time and writes clean code"
+- "Previously I thought all engineers were equal, but now I feel that experience and track record really matter"
+- "I believe reliability is best measured by consistent output over time"
+- "I've come to believe that track records are more important than potential"
+
+WRONG Examples (✗ THIRD-PERSON - DO NOT USE):
+- "The speaker thinks Alice is more reliable"
+- "They believe reliability matters"
+- "It is believed that Alice is better"
+
+If no genuine opinions are expressed (e.g., the response just says "I don't know"), return an empty list."""
+
+    try:
+        result = await llm_config.call(
+            messages=[
+                {"role": "system", "content": "You are converting opinions from text into first-person statements. Always use 'I think', 'I believe', 'I feel', etc. NEVER use third-person like 'The speaker' or 'They'."},
+                {"role": "user", "content": extraction_prompt}
+            ],
+            response_format=OpinionExtractionResponse,
+            scope="memory_extract_opinion"
+        )
+
+        # Format opinions with confidence score and convert to first-person
+        formatted_opinions = []
+        for op in result.opinions:
+            # Convert third-person to first-person if needed
+            opinion_text = op.opinion
+
+            # Replace common third-person patterns with first-person
+            def singularize_verb(verb):
+                if verb.endswith('es'):
+                    return verb[:-1]  # believes -> believe
+                elif verb.endswith('s'):
+                    return verb[:-1]  # thinks -> think
+                return verb
+
+            # Pattern: "The speaker/user [verb]..." -> "I [verb]..."
+            match = re.match(r'^(The speaker|The user|They|It is believed) (believes?|thinks?|feels?|says|asserts?|considers?)(\s+that)?(.*)$', opinion_text, re.IGNORECASE)
+            if match:
+                verb = singularize_verb(match.group(2))
+                that_part = match.group(3) or ""  # Keep " that" if present
+                rest = match.group(4)
+                opinion_text = f"I {verb}{that_part}{rest}"
+
+            # If still doesn't start with first-person, prepend "I believe that "
+            first_person_starters = ["I think", "I believe", "I feel", "In my view", "I've come to believe", "Previously I"]
+            if not any(opinion_text.startswith(starter) for starter in first_person_starters):
+                opinion_text = "I believe that " + opinion_text[0].lower() + opinion_text[1:]
+
+            formatted_opinions.append({
+                "text": opinion_text,
+                "confidence": op.confidence
+            })
+
+        return formatted_opinions
+
+    except Exception as e:
+        logger.warning(f"Failed to extract opinions: {str(e)}")
+        return []

hindsight-api/hindsight_api/migrations.py

@@ -37,14 +37,14 @@ def run_migrations(database_url: str, script_location: Optional[str] = None) ->
     Args:
         database_url: SQLAlchemy database URL (e.g., "postgresql://user:pass@host/db")
         script_location: Path to alembic migrations directory (e.g., "/path/to/alembic").
-                        If None, defaults to memora/alembic directory.
+                        If None, defaults to hindsight-api/alembic directory.
 
     Raises:
         RuntimeError: If migrations fail to complete
         FileNotFoundError: If script_location doesn't exist
 
     Example:
-        # Using default location (memora package)
+        # Using default location (hindsight_api package)
         run_migrations("postgresql://user:pass@host/db")
 
         # Using custom location (when importing from another project)
@@ -56,9 +56,9 @@ def run_migrations(database_url: str, script_location: Optional[str] = None) ->
     try:
         # Determine script location
         if script_location is None:
-            # Default: use the alembic directory in the memora package
-            # This file is in: memora/memora/migrations.py
-            # Default location is: memora/alembic
+            # Default: use the alembic directory in the hindsight_api package
+            # This file is in: hindsight-api/hindsight_api/migrations.py
+            # Default location is: hindsight-api/alembic
             package_root = Path(__file__).parent.parent
             script_location = str(package_root / "alembic")
 
@@ -86,6 +86,9 @@ def run_migrations(database_url: str, script_location: Optional[str] = None) ->
         # Uses Python's logging system instead of alembic.ini
         alembic_cfg.set_main_option("prepend_sys_path", ".")
 
+        # Set path_separator to avoid deprecation warning
+        alembic_cfg.set_main_option("path_separator", "os")
+
         # Run migrations to head (latest version)
         # Note: Alembic may call sys.exit() on errors instead of raising exceptions
         # We rely on the outer try/except and logging to catch issues
@@ -110,7 +113,7 @@ def check_migration_status(database_url: Optional[str] = None, script_location:
     Check current database schema version and latest available version.
 
     Args:
-        database_url: SQLAlchemy database URL. If None, uses MEMORA_API_DATABASE_URL env var.
+        database_url: SQLAlchemy database URL. If None, uses HINDSIGHT_API_DATABASE_URL env var.
         script_location: Path to alembic migrations directory. If None, uses default location.
 
     Returns:
@@ -124,9 +127,9 @@ def check_migration_status(database_url: Optional[str] = None, script_location:
 
         # Get database URL
         if database_url is None:
-            database_url = os.getenv("MEMORA_API_DATABASE_URL")
+            database_url = os.getenv("HINDSIGHT_API_DATABASE_URL")
         if not database_url:
-            logger.warning("Database URL not provided and MEMORA_API_DATABASE_URL not set, cannot check migration status")
+            logger.warning("Database URL not provided and HINDSIGHT_API_DATABASE_URL not set, cannot check migration status")
             return None, None
 
         # Get current revision from database
@@ -148,6 +151,7 @@ def check_migration_status(database_url: Optional[str] = None, script_location:
         # Create config programmatically
         alembic_cfg = Config()
         alembic_cfg.set_main_option("script_location", script_location)
+        alembic_cfg.set_main_option("path_separator", "os")
 
         script = ScriptDirectory.from_config(alembic_cfg)
         head_rev = script.get_current_head()

hindsight-api/hindsight_api/pg0.py

@@ -0,0 +1,416 @@
+import asyncio
+import json
+import logging
+import os
+import platform
+import shutil
+import stat
+import subprocess
+import sys
+from pathlib import Path
+from typing import Optional
+
+import httpx
+
+logger = logging.getLogger(__name__)
+
+DEFAULT_DATA_DIR = Path(os.environ.get("HINDSIGHT_API_PG0_DATA_DIR", Path.home() / ".hindsight" / "pg_data"))
+DEFAULT_INSTALL_DIR = Path.home() / ".hindsight" / "bin"
+BINARY_NAME = "pg0"
+DEFAULT_PORT = 5555
+DEFAULT_USERNAME = "hindsight"
+DEFAULT_PASSWORD = "hindsight"
+DEFAULT_DATABASE = "hindsight"
+
+
+def get_platform_binary_name() -> str:
+    """Get the appropriate binary name for the current platform.
+
+    Supported platforms:
+    - macOS ARM64 (darwin-aarch64)
+    - Linux x86_64
+    - Windows x86_64
+    """
+    system = platform.system().lower()
+    machine = platform.machine().lower()
+
+    # Normalize architecture names
+    if machine in ("x86_64", "amd64"):
+        arch = "x86_64"
+    elif machine in ("arm64", "aarch64"):
+        arch = "aarch64"
+    else:
+        raise RuntimeError(
+            f"Embedded PostgreSQL is not supported on architecture: {machine}. "
+            f"Supported architectures: x86_64/amd64 (Linux, Windows), aarch64/arm64 (macOS)"
+        )
+
+    if system == "darwin" and arch == "aarch64":
+        return "pg0-darwin-aarch64"
+    elif system == "linux" and arch == "x86_64":
+        return "pg0-linux-x86_64"
+    elif system == "windows" and arch == "x86_64":
+        return "pg0-windows-x86_64.exe"
+    else:
+        raise RuntimeError(
+            f"Embedded PostgreSQL is not supported on {system}-{arch}. "
+            f"Supported platforms: darwin-aarch64 (macOS ARM), linux-x86_64, windows-x86_64"
+        )
+
+
+def get_download_url(
+    version: str = "latest",
+    repo: str = "vectorize-io/pg0",
+) -> str:
+    """
+    """
+    # Check for direct URL override
+    binary_name = get_platform_binary_name()
+
+    if version == "latest":
+        return f"https://github.com/{repo}/releases/latest/download/{binary_name}"
+    else:
+        return f"https://github.com/{repo}/releases/download/{version}/{binary_name}"
+
+
+class EmbeddedPostgres:
+    """
+    Manages an embedded PostgreSQL server instance.
+
+    This class handles:
+    - Downloading and installing the embedded-postgres CLI
+    - Starting/stopping the PostgreSQL server
+    - Getting the connection URI
+
+    Example:
+        pg = EmbeddedPostgres(data_dir="~/.myapp/data")
+        await pg.ensure_installed()
+        await pg.start()
+        uri = await pg.get_uri()
+        # ... use uri with asyncpg ...
+        await pg.stop()
+    """
+
+    def __init__(
+        self,
+        data_dir: Optional[Path] = None,
+        install_dir: Optional[Path] = None,
+        version: str = "latest",
+        port: int = DEFAULT_PORT,
+        username: str = DEFAULT_USERNAME,
+        password: str = DEFAULT_PASSWORD,
+        database: str = DEFAULT_DATABASE,
+    ):
+        """
+        Initialize the embedded PostgreSQL manager.
+
+        Args:
+            data_dir: Directory to store PostgreSQL data. Defaults to ~/.hindsight/pg_data
+            install_dir: Directory to install the CLI binary. Defaults to ~/.hindsight/bin
+            version: Version of embedded-postgres to use. Defaults to "latest"
+            port: Port to listen on. Defaults to 5555
+            username: Username for the database. Defaults to "hindsight"
+            password: Password for the database. Defaults to "hindsight"
+            database: Database name to create. Defaults to "hindsight"
+        """
+        self.data_dir = Path(data_dir or DEFAULT_DATA_DIR).expanduser()
+        self.install_dir = Path(install_dir or DEFAULT_INSTALL_DIR).expanduser()
+        self.version = version
+        self.port = port
+        self.username = username
+        self.password = password
+        self.database = database
+
+        # Binary path
+        binary_name = "pg0.exe" if platform.system() == "Windows" else "pg0"
+        self.binary_path = self.install_dir / binary_name
+
+        self._process: Optional[subprocess.Popen] = None
+
+    def is_installed(self) -> bool:
+        """Check if the embedded-postgres CLI is installed."""
+        return self.binary_path.exists() and os.access(self.binary_path, os.X_OK)
+
+    async def ensure_installed(self) -> None:
+        """
+        Ensure the embedded-postgres CLI is installed.
+
+        Downloads and installs the binary if not already present.
+        """
+        if self.is_installed():
+            logger.debug(f"pg0 already installed at {self.binary_path}")
+            return
+
+        logger.info("Installing pg0 CLI...")
+
+        # Create install directory
+        self.install_dir.mkdir(parents=True, exist_ok=True)
+
+        # Download the binary
+        download_url = get_download_url(self.version)
+        logger.info(f"Downloading from {download_url}")
+
+        try:
+            async with httpx.AsyncClient(follow_redirects=True, timeout=300.0) as client:
+                response = await client.get(download_url)
+                response.raise_for_status()
+
+                # Write binary to disk
+                with open(self.binary_path, "wb") as f:
+                    f.write(response.content)
+
+                # Make executable on Unix
+                if platform.system() != "Windows":
+                    st = os.stat(self.binary_path)
+                    os.chmod(self.binary_path, st.st_mode | stat.S_IEXEC)
+
+                logger.info(f"Installed pg0 to {self.binary_path}")
+
+        except httpx.HTTPError as e:
+            raise RuntimeError(f"Failed to download pg0: {e}") from e
+
+    def _run_command(self, *args: str, capture_output: bool = True) -> subprocess.CompletedProcess:
+        """Run an embedded-postgres command synchronously."""
+        cmd = [str(self.binary_path), *args]
+
+        return subprocess.run(
+            cmd,
+            capture_output=capture_output,
+            text=True,
+        )
+
+    async def _run_command_async(self, *args: str) -> tuple[int, str, str]:
+        """Run an embedded-postgres command asynchronously."""
+        cmd = [str(self.binary_path), *args]
+
+        process = await asyncio.create_subprocess_exec(
+            *cmd,
+            stdout=asyncio.subprocess.PIPE,
+            stderr=asyncio.subprocess.PIPE,
+        )
+
+        stdout, stderr = await process.communicate()
+        return process.returncode, stdout.decode(), stderr.decode()
+
+    async def start(self) -> str:
+        """
+        Start the PostgreSQL server.
+
+        Returns:
+            The connection URI for the started server.
+
+        Raises:
+            RuntimeError: If the server fails to start.
+        """
+        if not self.is_installed():
+            raise RuntimeError("pg0 is not installed. Call ensure_installed() first.")
+
+        # Create data directory
+        self.data_dir.mkdir(parents=True, exist_ok=True)
+
+        logger.info(f"Starting embedded PostgreSQL (data: {self.data_dir}, port: {self.port})...")
+
+        returncode, stdout, stderr = await self._run_command_async(
+            "start",
+            "--port", str(self.port),
+            "--username", self.username,
+            "--password", self.password,
+            "--database", self.database,
+            "--data-dir", self.data_dir.as_posix()
+        )
+
+        if returncode != 0:
+            raise RuntimeError(f"Failed to start PostgreSQL: {stderr}")
+
+        logger.info("Embedded PostgreSQL started")
+
+        # Get and return the URI
+        return await self.get_uri()
+
+    async def stop(self) -> None:
+        """
+        Stop the PostgreSQL server.
+
+        Raises:
+            RuntimeError: If the server fails to stop.
+        """
+        if not self.is_installed():
+            return
+
+        logger.info("Stopping embedded PostgreSQL...")
+
+        returncode, stdout, stderr = await self._run_command_async("stop")
+
+        if returncode != 0:
+            # Don't raise if server wasn't running
+            if "not running" in stderr.lower():
+                logger.debug("PostgreSQL was not running")
+                return
+            raise RuntimeError(f"Failed to stop PostgreSQL: {stderr}")
+
+        logger.info("Embedded PostgreSQL stopped")
+
+    async def _get_info(self) -> dict:
+        """
+        Get info from pg0 using the `info -o json` command.
+
+        Returns:
+            Dictionary with 'running' (bool) and 'uri' (str) keys.
+
+        Raises:
+            RuntimeError: If unable to get info.
+        """
+        if not self.is_installed():
+            raise RuntimeError("pg0 is not installed.")
+
+        returncode, stdout, stderr = await self._run_command_async(
+            "info", "-o", "json")
+
+        if returncode != 0:
+            raise RuntimeError(f"Failed to get PostgreSQL info: {stderr}")
+
+        try:
+            return json.loads(stdout.strip())
+        except json.JSONDecodeError as e:
+            raise RuntimeError(f"Failed to parse pg0 info output: {e}")
+
+    async def get_uri(self) -> str:
+        """
+        Get the connection URI for the PostgreSQL server.
+
+        Returns:
+            PostgreSQL connection URI (e.g., postgresql://user:pass@localhost:5432/db)
+
+        Raises:
+            RuntimeError: If unable to get the URI or server is not running.
+        """
+        info = await self._get_info()
+        uri = info.get("uri")
+        if not uri:
+            raise RuntimeError("PostgreSQL server is not running or URI not available")
+        return uri
+
+    async def status(self) -> dict:
+        """
+        Get the status of the PostgreSQL server.
+
+        Returns:
+            Dictionary with status information including 'running' boolean and 'uri'.
+        """
+        if not self.is_installed():
+            return {"installed": False, "running": False}
+
+        try:
+            info = await self._get_info()
+            return {
+                "installed": True,
+                "running": info.get("running", False),
+                "uri": info.get("uri"),
+                "data_dir": str(self.data_dir),
+                "binary_path": str(self.binary_path),
+            }
+        except RuntimeError:
+            return {
+                "installed": True,
+                "running": False,
+                "data_dir": str(self.data_dir),
+                "binary_path": str(self.binary_path),
+            }
+
+    async def is_running(self) -> bool:
+        """Check if the PostgreSQL server is currently running."""
+        if not self.is_installed():
+            return False
+        try:
+            info = await self._get_info()
+            return info.get("running", False)
+        except RuntimeError:
+            return False
+
+    async def ensure_running(self) -> str:
+        """
+        Ensure the PostgreSQL server is running.
+
+        Installs if needed, starts if not running.
+
+        Returns:
+            The connection URI.
+        """
+        await self.ensure_installed()
+
+        if await self.is_running():
+            return await self.get_uri()
+
+        return await self.start()
+
+    def uninstall(self) -> None:
+        """Remove the embedded-postgres binary."""
+        if self.binary_path.exists():
+            self.binary_path.unlink()
+            logger.info(f"Removed {self.binary_path}")
+
+    def clear_data(self) -> None:
+        """Remove all PostgreSQL data (destructive!)."""
+        if self.data_dir.exists():
+            shutil.rmtree(self.data_dir)
+            logger.info(f"Removed data directory {self.data_dir}")
+
+
+# Convenience functions for simple usage
+
+_default_instance: Optional[EmbeddedPostgres] = None
+
+
+def get_embedded_postgres(
+    data_dir: Optional[Path] = None,
+    install_dir: Optional[Path] = None,
+) -> EmbeddedPostgres:
+    """
+    Get or create the default EmbeddedPostgres instance.
+
+    Args:
+        data_dir: Override default data directory
+        install_dir: Override default install directory
+
+    Returns:
+        EmbeddedPostgres instance
+    """
+    global _default_instance
+
+    if _default_instance is None or data_dir or install_dir:
+        _default_instance = EmbeddedPostgres(
+            data_dir=data_dir,
+            install_dir=install_dir,
+        )
+
+    return _default_instance
+
+
+async def start_embedded_postgres(
+    data_dir: Optional[Path] = None,
+) -> str:
+    """
+    Quick start function for embedded PostgreSQL.
+
+    Downloads, installs, and starts PostgreSQL in one call.
+
+    Args:
+        data_dir: Directory to store PostgreSQL data
+
+    Returns:
+        Connection URI string
+
+    Example:
+        db_url = await start_embedded_postgres()
+        conn = await asyncpg.connect(db_url)
+    """
+    pg = get_embedded_postgres(data_dir=data_dir)
+    return await pg.ensure_running()
+
+
+async def stop_embedded_postgres() -> None:
+    """Stop the default embedded PostgreSQL instance."""
+    global _default_instance
+
+    if _default_instance:
+        await _default_instance.stop()

hindsight-api/hindsight_api/web/__init__.py

@@ -3,10 +3,10 @@ Web interface for memory system.
 
 Provides FastAPI app and visualization interface.
 """
-from memora.api import create_app
+from hindsight_api.api import create_app
 
 # Note: Don't import app from .server here to avoid circular import warnings
-# when running with `python -m memora.web.server`
-# If you need the app, import it directly: from memora.web.server import app
+# when running with `python -m hindsight_api.web.server`
+# If you need the app, import it directly: from hindsight_api.web.server import app
 
 __all__ = ["create_app"]

hindsight-api/hindsight_api/web/server.py

@@ -4,27 +4,68 @@ FastAPI server for memory graph visualization and API.
 Provides REST API endpoints for memory operations and serves
 the interactive visualization interface.
 """
+import warnings
+
+# Filter deprecation warnings from third-party libraries
+warnings.filterwarnings("ignore", message="websockets.legacy is deprecated")
+warnings.filterwarnings("ignore", message="websockets.server.WebSocketServerProtocol is deprecated")
+
+import asyncio
+import atexit
 import logging
 import os
 import argparse
+import signal
+import sys
 
-from memora import TemporalSemanticMemory
-from memora.api import create_app
+from hindsight_api import MemoryEngine
+from hindsight_api.api import create_app
 
 # Disable tokenizers parallelism to avoid warnings
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
 
+
+def _cleanup_pg0():
+    """Synchronous cleanup function to stop pg0 on exit."""
+    global _memory
+    if _memory is not None and _memory._pg0 is not None:
+        try:
+            # Run async stop in a new event loop
+            loop = asyncio.new_event_loop()
+            loop.run_until_complete(_memory._pg0.stop())
+            loop.close()
+            print("\npg0 stopped.")
+        except Exception as e:
+            print(f"\nError stopping pg0: {e}")
+
+
+# Register cleanup on normal exit
+atexit.register(_cleanup_pg0)
+
+
+def _signal_handler(signum, frame):
+    """Handle SIGINT/SIGTERM to ensure pg0 cleanup."""
+    print(f"\nReceived signal {signum}, shutting down...")
+    _cleanup_pg0()
+    sys.exit(0)
+
+
+# Register signal handlers for graceful shutdown
+signal.signal(signal.SIGINT, _signal_handler)
+signal.signal(signal.SIGTERM, _signal_handler)
+
+
 # Create app at module level (required for uvicorn import string)
-_memory = TemporalSemanticMemory(
-    db_url=os.getenv("MEMORA_API_DATABASE_URL"),
-    memory_llm_provider=os.getenv("MEMORA_API_LLM_PROVIDER", "groq"),
-    memory_llm_api_key=os.getenv("MEMORA_API_LLM_API_KEY"),
-    memory_llm_model=os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-120b"),
-    memory_llm_base_url=os.getenv("MEMORA_API_LLM_BASE_URL") or None,
+_memory = MemoryEngine(
+    db_url=os.getenv("HINDSIGHT_API_DATABASE_URL", "pg0"),
+    memory_llm_provider=os.getenv("HINDSIGHT_API_LLM_PROVIDER", "groq"),
+    memory_llm_api_key=os.getenv("HINDSIGHT_API_LLM_API_KEY"),
+    memory_llm_model=os.getenv("HINDSIGHT_API_LLM_MODEL", "openai/gpt-oss-120b"),
+    memory_llm_base_url=os.getenv("HINDSIGHT_API_LLM_BASE_URL") or None,
 )
 
 # Check if MCP should be enabled
-mcp_enabled = os.getenv("MEMORA_API_MCP_ENABLED", "true").lower() == "true"
+mcp_enabled = os.getenv("HINDSIGHT_API_MCP_ENABLED", "true").lower() == "true"
 
 # Create unified app with both HTTP and optionally MCP
 app = create_app(
@@ -43,7 +84,7 @@ if __name__ == "__main__":
     # Parse CLI arguments
     parser = argparse.ArgumentParser(description="Memory Graph API Server")
     parser.add_argument("--host", default="0.0.0.0", help="Host to bind to (default: 0.0.0.0)")
-    parser.add_argument("--port", type=int, default=8080, help="Port to bind to (default: 8080)")
+    parser.add_argument("--port", type=int, default=8888, help="Port to bind to (default: 8888)")
     parser.add_argument("--reload", action="store_true", help="Enable auto-reload on code changes")
     parser.add_argument("--workers", type=int, default=1, help="Number of worker processes (default: 1)")
     parser.add_argument("--log-level", default="info", choices=["critical", "error", "warning", "info", "debug", "trace"],
@@ -58,7 +99,7 @@ if __name__ == "__main__":
 
     args = parser.parse_args()
 
-    app_ref = "memora.web.server:app"
+    app_ref = "hindsight_api.web.server:app"
 
     # Prepare uvicorn config
     uvicorn_config = {

hindsight-api/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["hatchling"]
 build-backend = "hatchling.build"
 
 [project]
-name = "memora"
+name = "hindsight-api"
 version = "0.0.7"
 description = "Temporal + Semantic + Entity Memory System for AI agents using PostgreSQL"
 readme = "README.md"
@@ -36,15 +36,24 @@ test = [
     "pytest>=7.0.0",
     "pytest-asyncio>=0.21.0",
     "pytest-timeout>=2.4.0",
+    "pytest-xdist>=3.0.0",
+    "filelock>=3.0.0",
+    "testcontainers[postgres]>=4.0.0",
 ]
 
 [tool.hatch.build.targets.wheel]
-packages = ["memora"]
+packages = ["hindsight_api"]
 
 [tool.pytest.ini_options]
 log_cli = true
 log_cli_level = "INFO"
 log_cli_format = "%(asctime)s %(levelname)s %(message)s"
 log_cli_date_format = "%Y-%m-%d %H:%M:%S"
-addopts = "--timeout 60 -p no:warnings"
+addopts = "--timeout 60 -n auto --durations=10 -v"
+asyncio_mode = "auto"
 asyncio_default_fixture_loop_scope = "function"
+log_auto_indent = true
+filterwarnings = [
+    "ignore:The @wait_container_is_ready decorator is deprecated:DeprecationWarning",
+    "ignore::RuntimeWarning:asyncio",
+]

hindsight-api/tests/conftest.py

@@ -0,0 +1,130 @@
+"""
+Pytest configuration and shared fixtures.
+"""
+import pytest
+import pytest_asyncio
+import os
+import filelock
+from pathlib import Path
+from dotenv import load_dotenv
+from hindsight_api import MemoryEngine, LLMConfig, SentenceTransformersEmbeddings
+import asyncpg
+from testcontainers.postgres import PostgresContainer
+
+from hindsight_api.engine.cross_encoder import SentenceTransformersCrossEncoder
+from hindsight_api.engine.query_analyzer import TransformerQueryAnalyzer
+
+
+# Load environment variables from .env at the start of test session
+def pytest_configure(config):
+    """Load environment variables before running tests."""
+    # Look for .env in the workspace root (two levels up from tests dir)
+    env_file = Path(__file__).parent.parent.parent / ".env"
+    if env_file.exists():
+        load_dotenv(env_file)
+    else:
+        print(f"Warning: {env_file} not found, tests may fail without proper configuration")
+
+
+@pytest.fixture(scope="session")
+def postgres_container(tmp_path_factory, worker_id):
+    """
+    Start a postgres container shared across all test workers.
+    Uses filelock to ensure only one worker starts the container.
+
+    - worker_id == "master": running without -n (single process)
+    - worker_id == "gw0", "gw1", etc.: running with -n (parallel workers)
+    """
+    # Get shared temp dir (same for all workers)
+    if worker_id == "master":
+        root_tmp_dir = tmp_path_factory.getbasetemp()
+    else:
+        root_tmp_dir = tmp_path_factory.getbasetemp().parent
+
+    db_url_file = root_tmp_dir / "postgres_url"
+    lock_file = root_tmp_dir / "postgres.lock"
+    container = None
+
+    with filelock.FileLock(str(lock_file)):
+        if db_url_file.exists():
+            # Another worker already started the container
+            db_url = db_url_file.read_text()
+        else:
+            # First worker - start the container
+            container = PostgresContainer("pgvector/pgvector:pg16")
+            container.start()
+            db_url = container.get_connection_url().replace("postgresql+psycopg2://", "postgresql://")
+            db_url_file.write_text(db_url)
+
+            # Run migrations
+            from hindsight_api.migrations import run_migrations
+            run_migrations(db_url)
+
+    os.environ["HINDSIGHT_API_DATABASE_URL"] = db_url
+    yield db_url
+
+    # Only the worker that started the container stops it
+    if container is not None:
+        container.stop()
+
+
+@pytest.fixture(scope="session")
+def llm_config():
+    """
+    Provide LLM configuration for tests.
+    This can be used by tests that need to call LLM directly without memory system.
+    """
+    return LLMConfig.for_memory()
+
+
+@pytest.fixture(scope="session")
+def embeddings():
+
+    return SentenceTransformersEmbeddings("BAAI/bge-small-en-v1.5")
+
+
+
+@pytest.fixture(scope="session")
+def cross_encoder():
+
+    return SentenceTransformersCrossEncoder()
+
+@pytest.fixture(scope="session")
+def query_analyzer():
+    return TransformerQueryAnalyzer()
+
+
+
+
+@pytest_asyncio.fixture(scope="function")
+async def memory(postgres_container, embeddings, cross_encoder, query_analyzer):
+    """
+    Provide a MemoryEngine instance for each test.
+
+    Must be function-scoped because:
+    1. pytest-xdist runs tests in separate processes with different event loops
+    2. asyncpg pools are bound to the event loop that created them
+    3. Each test needs its own pool in its own event loop
+
+    Uses small pool sizes since tests run in parallel and share a single
+    testcontainer PostgreSQL instance with limited resources.
+    """
+    mem = MemoryEngine(
+        db_url=postgres_container,
+        memory_llm_provider=os.getenv("HINDSIGHT_API_LLM_PROVIDER", "groq"),
+        memory_llm_api_key=os.getenv("HINDSIGHT_API_LLM_API_KEY"),
+        memory_llm_model=os.getenv("HINDSIGHT_API_LLM_MODEL", "openai/gpt-oss-120b"),
+        memory_llm_base_url=os.getenv("HINDSIGHT_API_LLM_BASE_URL") or None,
+        embeddings=embeddings,
+        cross_encoder=cross_encoder,
+        query_analyzer=query_analyzer,
+        pool_min_size=1,
+        pool_max_size=5,
+    )
+    await mem.initialize()
+    yield mem
+    try:
+        if mem._pool and not mem._pool._closing:
+            await mem.close()
+    except Exception:
+        pass

hindsight-api/tests/test_agents_api.py

@@ -0,0 +1,254 @@
+"""
+Tests for agent management API (profile, personality, background).
+"""
+import pytest
+import uuid
+from hindsight_api import MemoryEngine
+from hindsight_api.api import CreateAgentRequest, PersonalityTraits
+
+
+def unique_agent_id(prefix: str) -> str:
+    """Generate a unique agent ID for testing."""
+    return f"{prefix}_{uuid.uuid4().hex[:8]}"
+
+
+class TestAgentProfile:
+    """Tests for agent profile management."""
+
+    @pytest.mark.asyncio
+    async def test_get_agent_profile_creates_default(self, memory: MemoryEngine):
+        """Test that getting a profile for a new agent creates default personality."""
+        agent_id = unique_agent_id("test_profile_default")
+
+        profile = await memory.get_agent_profile(agent_id)
+
+        assert profile is not None
+        assert "personality" in profile
+        assert "background" in profile
+
+        personality = profile["personality"]
+        assert personality["openness"] == 0.5
+        assert personality["conscientiousness"] == 0.5
+        assert personality["extraversion"] == 0.5
+        assert personality["agreeableness"] == 0.5
+        assert personality["neuroticism"] == 0.5
+        assert personality["bias_strength"] == 0.5
+
+        assert profile["background"] == ""
+
+    @pytest.mark.asyncio
+    async def test_update_agent_personality(self, memory: MemoryEngine):
+        """Test updating agent personality traits."""
+        agent_id = unique_agent_id("test_profile_update")
+
+        profile = await memory.get_agent_profile(agent_id)
+        assert profile["personality"]["openness"] == 0.5
+
+        new_personality = {
+            "openness": 0.8,
+            "conscientiousness": 0.6,
+            "extraversion": 0.7,
+            "agreeableness": 0.4,
+            "neuroticism": 0.3,
+            "bias_strength": 0.9,
+        }
+        await memory.update_agent_personality(agent_id, new_personality)
+
+        updated_profile = await memory.get_agent_profile(agent_id)
+        for key in new_personality:
+            assert abs(updated_profile["personality"][key] - new_personality[key]) < 0.001
+
+    @pytest.mark.asyncio
+    async def test_list_agents(self, memory: MemoryEngine):
+        """Test listing all agents."""
+        agent_id_1 = unique_agent_id("test_list")
+        agent_id_2 = unique_agent_id("test_list")
+        agent_id_3 = unique_agent_id("test_list")
+
+        await memory.get_agent_profile(agent_id_1)
+        await memory.get_agent_profile(agent_id_2)
+        await memory.get_agent_profile(agent_id_3)
+
+        agents = await memory.list_agents()
+
+        agent_ids = [a["agent_id"] for a in agents]
+        assert agent_id_1 in agent_ids
+        assert agent_id_2 in agent_ids
+        assert agent_id_3 in agent_ids
+
+        for agent in agents:
+            assert "agent_id" in agent
+            assert "personality" in agent
+            assert "background" in agent
+            assert "created_at" in agent
+            assert "updated_at" in agent
+
+
+class TestAgentBackground:
+    """Tests for agent background management."""
+
+    @pytest.mark.asyncio
+    async def test_merge_agent_background(self, memory: MemoryEngine):
+        """Test merging agent background information."""
+        agent_id = unique_agent_id("test_profile_merge")
+
+        profile = await memory.get_agent_profile(agent_id)
+        assert profile["background"] == ""
+
+        result1 = await memory.merge_agent_background(
+            agent_id,
+            "I was born in Texas",
+            update_personality=False
+        )
+        assert "Texas" in result1["background"]
+
+        result2 = await memory.merge_agent_background(
+            agent_id,
+            "I have 10 years of startup experience",
+            update_personality=False
+        )
+        assert "Texas" in result2["background"] or "startup" in result2["background"]
+
+        final_profile = await memory.get_agent_profile(agent_id)
+        assert final_profile["background"] != ""
+
+    @pytest.mark.asyncio
+    async def test_merge_background_handles_conflicts(self, memory: MemoryEngine):
+        """Test that merging background handles conflicts (new overwrites old)."""
+        agent_id = unique_agent_id("test_profile_conflict")
+
+        result1 = await memory.merge_agent_background(
+            agent_id,
+            "I was born in Colorado",
+            update_personality=False
+        )
+        assert "Colorado" in result1["background"]
+
+        result2 = await memory.merge_agent_background(
+            agent_id,
+            "You were born in Texas",
+            update_personality=False
+        )
+        assert "Texas" in result2["background"]
+
+
+class TestAgentEndpoint:
+    """Tests for agent PUT endpoint logic."""
+
+    @pytest.mark.asyncio
+    async def test_put_agent_create(self, memory: MemoryEngine):
+        """Test creating an agent via PUT endpoint."""
+        agent_id = unique_agent_id("test_put_create")
+
+        request = CreateAgentRequest(
+            personality=PersonalityTraits(
+                openness=0.8,
+                conscientiousness=0.6,
+                extraversion=0.5,
+                agreeableness=0.7,
+                neuroticism=0.3,
+                bias_strength=0.7
+            ),
+            background="I am a creative software engineer"
+        )
+
+        profile = await memory.get_agent_profile(agent_id)
+
+        if request.personality is not None:
+            await memory.update_agent_personality(
+                agent_id,
+                request.personality.model_dump()
+            )
+
+        if request.background is not None:
+            pool = await memory._get_pool()
+            async with pool.acquire() as conn:
+                await conn.execute(
+                    """
+                    UPDATE agents
+                    SET background = $2,
+                        updated_at = NOW()
+                    WHERE agent_id = $1
+                    """,
+                    agent_id,
+                    request.background
+                )
+
+        final_profile = await memory.get_agent_profile(agent_id)
+
+        assert final_profile["personality"]["openness"] == 0.8
+        assert final_profile["personality"]["bias_strength"] == 0.7
+        assert final_profile["background"] == "I am a creative software engineer"
+
+    @pytest.mark.asyncio
+    async def test_put_agent_partial_update(self, memory: MemoryEngine):
+        """Test updating only background."""
+        agent_id = unique_agent_id("test_put_partial")
+
+        request = CreateAgentRequest(
+            background="I am a data scientist"
+        )
+
+        profile = await memory.get_agent_profile(agent_id)
+
+        if request.background is not None:
+            pool = await memory._get_pool()
+            async with pool.acquire() as conn:
+                await conn.execute(
+                    """
+                    UPDATE agents
+                    SET background = $2,
+                        updated_at = NOW()
+                    WHERE agent_id = $1
+                    """,
+                    agent_id,
+                    request.background
+                )
+
+        final_profile = await memory.get_agent_profile(agent_id)
+
+        assert final_profile["personality"]["openness"] == 0.5
+        assert final_profile["background"] == "I am a data scientist"
+
+
+class TestAgentPersonalityIntegration:
+    """Tests for personality integration with other features."""
+
+    @pytest.mark.asyncio
+    async def test_think_uses_personality(self, memory: MemoryEngine):
+        """Test that THINK operation uses agent personality."""
+        agent_id = unique_agent_id("test_think")
+
+        personality = {
+            "openness": 0.9,
+            "conscientiousness": 0.2,
+            "extraversion": 0.8,
+            "agreeableness": 0.1,
+            "neuroticism": 0.7,
+            "bias_strength": 0.9,
+        }
+        await memory.update_agent_personality(agent_id, personality)
+
+        await memory.merge_agent_background(
+            agent_id,
+            "I am a creative artist who values innovation over tradition",
+            update_personality=False
+        )
+
+        await memory.put_batch_async(
+            agent_id=agent_id,
+            contents=[
+                {"content": "Traditional painting techniques have been used for centuries"},
+                {"content": "Modern digital art is changing the art world"}
+            ],
+            document_id="art_facts"
+        )
+
+        result = await memory.think_async(
+            agent_id=agent_id,
+            query="What do you think about traditional vs modern art?",
+            thinking_budget=50
+        )
+
+        assert result.text is not None
+        assert len(result.text) > 0

hindsight-api/tests/test_batch_chunking.py

@@ -0,0 +1,58 @@
+"""Test automatic batch chunking based on character count."""
+import asyncio
+import pytest
+from hindsight_api import MemoryEngine
+import os
+
+
+@pytest.mark.asyncio
+async def test_large_batch_auto_chunks(memory):
+    agent_id = "test_chunking_agent"
+    # Create a large batch that should trigger chunking
+    # Each item is ~2000 chars, so 30 items = 60k chars (exceeds 50k threshold)
+    large_content = "Alice met with Bob at the coffee shop. " * 50  # ~2000 chars
+    contents = [
+        {"content": large_content, "context": f"conversation_{i}"}
+        for i in range(30)
+    ]
+
+    # Calculate total chars
+    total_chars = sum(len(item["content"]) for item in contents)
+    print(f"\nTotal characters: {total_chars:,}")
+    print(f"Should trigger chunking: {total_chars > 50_000}")
+
+    # Ingest the large batch (should auto-chunk)
+    result = await memory.put_batch_async(
+        agent_id=agent_id,
+        contents=contents
+    )
+
+    # Verify we got results back
+    assert len(result) == 30, f"Expected 30 results, got {len(result)}"
+    print(f"Successfully ingested {len(result)} items (auto-chunked)")
+
+
+@pytest.mark.asyncio
+async def test_small_batch_no_chunking(memory):
+    agent_id = "test_no_chunking_agent"
+
+    # Create a small batch that should NOT trigger chunking
+    contents = [
+        {"content": "Alice works at Google", "context": "conversation_1"},
+        {"content": "Bob loves Python", "context": "conversation_2"}
+    ]
+
+    # Calculate total chars
+    total_chars = sum(len(item["content"]) for item in contents)
+    print(f"\nTotal characters: {total_chars:,}")
+    print(f"Should NOT trigger chunking: {total_chars <= 50_000}")
+
+    # Ingest the small batch (should NOT auto-chunk)
+    result = await memory.put_batch_async(
+        agent_id=agent_id,
+        contents=contents
+    )
+
+    # Verify we got results back
+    assert len(result) == 2, f"Expected 2 results, got {len(result)}"
+    print(f"Successfully ingested {len(result)} items (no chunking)")

hindsight-api/tests/test_chunking.py

@@ -2,7 +2,7 @@
 Test chunking functionality for large documents.
 """
 import pytest
-from memora.fact_extraction import chunk_text
+from hindsight_api.engine.fact_extraction import chunk_text
 
 
 def test_chunk_text_small():
@@ -58,6 +58,3 @@ def test_chunk_text_64k():
     combined_length = sum(len(chunk) for chunk in chunks)
     assert combined_length >= len(text) * 0.95, "Lost too much content during chunking"
 
-
-if __name__ == "__main__":
-    pytest.main([__file__, "-v"])

hindsight-api/tests/test_fact_ordering.py

@@ -7,31 +7,14 @@ distinguish between things said earlier vs later.
 """
 import pytest
 from datetime import datetime, timezone
-from memora import TemporalSemanticMemory
+from hindsight_api import MemoryEngine
 import os
 
 
 @pytest.mark.asyncio
-async def test_fact_ordering_within_conversation():
-    """
-    Test that facts extracted from one conversation get incremental time offsets
-    to preserve their ordering for retrieval.
-    """
-
-    # Create memory instance
-    memory = TemporalSemanticMemory(
-        db_url=os.getenv("MEMORA_API_DATABASE_URL"),
-        memory_llm_provider=os.getenv("MEMORA_API_LLM_PROVIDER", "groq"),
-        memory_llm_api_key=os.getenv("MEMORA_API_LLM_API_KEY"),
-        memory_llm_model=os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-20b"),
-    )
-    await memory.initialize()
-
+async def test_fact_ordering_within_conversation(memory):
     agent_id = "test_ordering_agent"
 
-    # Clear any existing data
-    await memory.delete_agent(agent_id)
-
     # Get/create agent (auto-creates with defaults)
     await memory.get_agent_profile(agent_id)
 
@@ -74,20 +57,20 @@ Marcus: Yeah, I realized I was being too optimistic about their defense.
         max_tokens=8192
     )
 
-    print(f"\n=== Retrieved {len(results['results'])} facts ===")
-    for i, result in enumerate(results['results']):
-        print(f"{i+1}. [{result['event_date']}] {result['text'][:100]}")
+    print(f"\n=== Retrieved {len(results.results)} facts ===")
+    for i, result in enumerate(results.results):
+        print(f"{i+1}. [{result.event_date}] {result.text[:100]}")
 
     # Get all agent facts (Marcus's statements)
-    agent_facts = [r for r in results['results'] if r.get('fact_type') == 'agent']
+    agent_facts = [r for r in results.results if r.fact_type == 'agent']
 
     print(f"\n=== Agent facts (Marcus's statements) ===")
     for i, fact in enumerate(agent_facts):
-        print(f"{i+1}. [{fact['event_date']}] {fact['text']}")
+        print(f"{i+1}. [{fact.event_date}] {fact.text}")
 
     # Check that agent facts have different timestamps
     if len(agent_facts) >= 2:
-        timestamps = [datetime.fromisoformat(f['event_date'].replace('Z', '+00:00')) for f in agent_facts]
+        timestamps = [datetime.fromisoformat(f.event_date.replace('Z', '+00:00')) for f in agent_facts]
 
         # Verify timestamps are different (have time offsets)
         unique_timestamps = set(timestamps)
@@ -111,7 +94,7 @@ Marcus: Yeah, I realized I was being too optimistic about their defense.
 
     # Verify that retrieval returns facts in chronological order
     # The first prediction should come before the changed prediction
-    agent_texts = [f['text'].lower() for f in agent_facts]
+    agent_texts = [f.text.lower() for f in agent_facts]
 
     # Look for evidence of the sequence
     has_first_prediction = any('27' in text and '24' in text for text in agent_texts)
@@ -122,8 +105,8 @@ Marcus: Yeah, I realized I was being too optimistic about their defense.
         first_idx = next(i for i, text in enumerate(agent_texts) if '27' in text and '24' in text)
         changed_idx = next(i for i, text in enumerate(agent_texts) if 'chang' in text or 'by 3' in text or 'realized' in text)
 
-        print(f"\nFirst prediction at index {first_idx}: {agent_facts[first_idx]['text'][:100]}")
-        print(f"Changed prediction at index {changed_idx}: {agent_facts[changed_idx]['text'][:100]}")
+        print(f"\nFirst prediction at index {first_idx}: {agent_facts[first_idx].text[:100]}")
+        print(f"Changed prediction at index {changed_idx}: {agent_facts[changed_idx].text[:100]}")
 
         # The original prediction should come before the changed one
         assert timestamps[first_idx] < timestamps[changed_idx], \
@@ -138,24 +121,10 @@ Marcus: Yeah, I realized I was being too optimistic about their defense.
 
 
 @pytest.mark.asyncio
-async def test_multiple_documents_ordering():
-    """
-    Test that facts from different documents get separate time offsets,
-    so facts within each document maintain their order.
-    """
-
-    memory = TemporalSemanticMemory(
-        db_url=os.getenv("MEMORA_API_DATABASE_URL"),
-        memory_llm_provider=os.getenv("MEMORA_API_LLM_PROVIDER", "groq"),
-        memory_llm_api_key=os.getenv("MEMORA_API_LLM_API_KEY"),
-        memory_llm_model=os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-20b"),
-    )
-    await memory.initialize()
+async def test_multiple_documents_ordering(memory):
 
     agent_id = "test_multi_doc_agent"
 
-    # Clear and create agent
-    await memory.delete_agent(agent_id)
     await memory.get_agent_profile(agent_id)  # Auto-creates with defaults
 
     # Two separate conversations with same base time
@@ -191,15 +160,15 @@ Alice: I reconsidered the team's experience level.
         max_tokens=8192
     )
 
-    print(f"\n=== Retrieved {len(results['results'])} agent facts ===")
-    agent_facts = [r for r in results['results'] if r.get('fact_type') == 'agent']
+    print(f"\n=== Retrieved {len(results.results)} agent facts ===")
+    agent_facts = [r for r in results.results if r.fact_type == 'agent']
 
     for i, fact in enumerate(agent_facts):
-        print(f"{i+1}. [{fact['event_date']}] {fact['text'][:80]}")
+        print(f"{i+1}. [{fact.event_date}] {fact.text[:80]}")
 
     # Each conversation's facts should have different timestamps
     if len(agent_facts) >= 2:
-        timestamps = [datetime.fromisoformat(f['event_date'].replace('Z', '+00:00')) for f in agent_facts]
+        timestamps = [datetime.fromisoformat(f.event_date.replace('Z', '+00:00')) for f in agent_facts]
         unique_timestamps = set(timestamps)
 
         assert len(unique_timestamps) >= 2, \

hindsight-api/tests/test_http_api_integration.py

@@ -1,5 +1,5 @@
 """
-Integration test for the complete Memora API.
+Integration test for the complete Hindsight API.
 
 Tests all endpoints by starting a FastAPI server and making HTTP requests.
 """
@@ -7,7 +7,7 @@ import pytest
 import pytest_asyncio
 import httpx
 from datetime import datetime
-from memora.api import create_app
+from hindsight_api.api import create_app
 
 
 @pytest_asyncio.fixture

hindsight-api/tests/test_mcp_api_integration.py

@@ -1,17 +1,42 @@
-"""Test MCP server with real server and client."""
+"""
+Integration test for the MCP (Model Context Protocol) server.
 
+Tests MCP endpoints by starting a FastAPI server with MCP enabled and using the MCP client.
+
+Note: MCP server is integrated with the web server. These tests require HINDSIGHT_API_MCP_ENABLED=true.
+"""
 import asyncio
-import os
 import pytest
+import pytest_asyncio
+import httpx
 from mcp import ClientSession
 from mcp.client.sse import sse_client
+from hindsight_api.api import create_app
 
 
-# Note: MCP server tests now require the full web server to be running
-# with MEMORA_API_MCP_ENABLED=true since there's no standalone MCP server anymore.
-# These tests are kept for documentation but may need manual server setup.
+@pytest_asyncio.fixture
+async def mcp_server(memory):
+    """Start the FastAPI app with MCP enabled and return the SSE URL."""
+    app = create_app(
+        memory,
+        run_migrations=False,
+        initialize_memory=False,
+        mcp_enabled=True,
+        default_agent_id="test_mcp_agent"
+    )
 
-pytest.skip("MCP server is now integrated with web server. Run web server with MEMORA_API_MCP_ENABLED=true to test.", allow_module_level=True)
+    # Use httpx to create a test server
+    transport = httpx.ASGITransport(app=app)
+    async with httpx.AsyncClient(transport=transport, base_url="http://test") as client:
+        # The MCP SSE endpoint is at /mcp/sse
+        # We need to yield the base URL for sse_client to connect
+        # However, sse_client expects a real URL, not a test client
+        # So we'll start a real server on a random port
+        pass
+
+    # For now, skip these tests as they require a real server
+    # The sse_client doesn't work with ASGI test transport
+    pytest.skip("MCP tests require a real running server. Run: HINDSIGHT_API_MCP_ENABLED=true uvicorn hindsight_api.api:app")
 
 
 @pytest.mark.asyncio
@@ -27,12 +52,12 @@ async def test_mcp_server_tools_via_sse(mcp_server):
             tools_list = await session.list_tools()
             print(f"Tools: {tools_list}")
             tool_names = [t.name for t in tools_list.tools]
-            assert "memora_search" in tool_names
-            assert "memora_put" in tool_names
+            assert "hindsight_search" in tool_names
+            assert "hindsight_put" in tool_names
 
-            # Test 2: Call memora_put
+            # Test 2: Call hindsight_put
             put_result = await session.call_tool(
-                "memora_put",
+                "hindsight_put",
                 arguments={
                     "content": "User loves Python programming",
                     "context": "programming_preferences",
@@ -45,9 +70,9 @@ async def test_mcp_server_tools_via_sse(mcp_server):
             # Wait a bit for indexing
             await asyncio.sleep(1)
 
-            # Test 3: Call memora_search
+            # Test 3: Call hindsight_search
             search_result = await session.call_tool(
-                "memora_search",
+                "hindsight_search",
                 arguments={
                     "query": "What programming languages does the user like?",
                     "max_tokens": 4096,
@@ -71,7 +96,7 @@ async def test_multiple_concurrent_requests(mcp_server):
             async def make_search(idx):
                 try:
                     result = await session.call_tool(
-                        "memora_search",
+                        "hindsight_search",
                         arguments={
                             "query": f"test query {idx}",
                             "explanation": f"Concurrent test {idx}"
@@ -120,7 +145,7 @@ async def test_race_condition_with_rapid_requests(mcp_server):
 
                     # Make request immediately after initialization
                     result = await session.call_tool(
-                        "memora_search",
+                        "hindsight_search",
                         arguments={
                             "query": f"rapid query {idx}",
                             "max_tokens": 2048

hindsight-api/tests/test_query_analyzer.py

@@ -3,16 +3,14 @@ Test query analyzer for temporal extraction.
 """
 import pytest
 from datetime import datetime
-from memora.query_analyzer import TransformerQueryAnalyzer, QueryAnalysis
+from hindsight_api.engine.query_analyzer import TransformerQueryAnalyzer, QueryAnalysis
 
 
-def test_query_analyzer_june_2024():
-    """Test extracting 'june 2024' from query."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_june_2024(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "june 2024"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -26,13 +24,11 @@ def test_query_analyzer_june_2024():
     assert analysis.temporal_constraint.end_date.day == 30
 
 
-def test_query_analyzer_dogs_june_2023():
-    """Test extracting temporal info from 'dogs in June 2023'."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_dogs_june_2023(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "dogs in June 2023"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -46,13 +42,11 @@ def test_query_analyzer_dogs_june_2023():
     assert analysis.temporal_constraint.end_date.day == 30
 
 
-def test_query_analyzer_march_2023():
-    """Test extracting 'March 2023' from query."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_march_2023(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "March 2023"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -66,13 +60,11 @@ def test_query_analyzer_march_2023():
     assert analysis.temporal_constraint.end_date.day == 31
 
 
-def test_query_analyzer_last_year():
-    """Test extracting 'last year' from query."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_last_year(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "last year"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -86,13 +78,11 @@ def test_query_analyzer_last_year():
     assert analysis.temporal_constraint.end_date.day == 31
 
 
-def test_query_analyzer_no_temporal():
-    """Test that queries without temporal info return None."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_no_temporal(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "what is the weather"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -100,13 +90,11 @@ def test_query_analyzer_no_temporal():
     assert analysis.temporal_constraint is None, "Should not extract temporal constraint"
 
 
-def test_query_analyzer_activities_june_2024():
-    """Test extracting temporal info from 'melanie activities in june 2024'."""
-    analyzer = TransformerQueryAnalyzer()
+def test_query_analyzer_activities_june_2024(query_analyzer):
     reference_date = datetime(2025, 1, 15, 12, 0, 0)
 
     query = "melanie activities in june 2024"
-    analysis = analyzer.analyze(query, reference_date)
+    analysis = query_analyzer.analyze(query, reference_date)
 
     print(f"\nQuery: '{query}'")
     print(f"Analysis: {analysis}")
@@ -120,5 +108,3 @@ def test_query_analyzer_activities_june_2024():
     assert analysis.temporal_constraint.end_date.day == 30
 
 
-if __name__ == "__main__":
-    pytest.main([__file__, "-v", "-s"])

hindsight-api/tests/test_search_trace.py

@@ -2,7 +2,7 @@
 Test search tracing functionality.
 """
 import pytest
-from memora.search_trace import SearchTrace
+from hindsight_api import SearchTrace
 from datetime import datetime, timezone
 
 

hindsight-api/tests/test_temporal_ranges.py

@@ -3,7 +3,7 @@ import asyncio
 import os
 from datetime import datetime, timezone, timedelta
 import pytest
-from memora import TemporalSemanticMemory
+from hindsight_api import MemoryEngine
 
 
 @pytest.mark.asyncio
@@ -11,11 +11,11 @@ async def test_temporal_ranges_are_written():
     """Test that occurred_start, occurred_end, and mentioned_at are actually written to database."""
 
     # Initialize memory system
-    memory = TemporalSemanticMemory(
-        db_url=os.getenv("MEMORA_API_DATABASE_URL", "postgresql://memora:memora_dev@localhost:5432/memora"),
-        memory_llm_provider=os.getenv("MEMORA_API_LLM_PROVIDER", "groq"),
-        memory_llm_api_key=os.getenv("MEMORA_API_LLM_API_KEY"),
-        memory_llm_model=os.getenv("MEMORA_API_LLM_MODEL", "openai/gpt-oss-20b"),
+    memory = MemoryEngine(
+        db_url=os.getenv("HINDSIGHT_API_DATABASE_URL", "postgresql://hindsight:hindsight_dev@localhost:5432/hindsight"),
+        memory_llm_provider=os.getenv("HINDSIGHT_API_LLM_PROVIDER", "groq"),
+        memory_llm_api_key=os.getenv("HINDSIGHT_API_LLM_API_KEY"),
+        memory_llm_model=os.getenv("HINDSIGHT_API_LLM_MODEL", "openai/gpt-oss-20b"),
     )
     await memory.initialize()
 

hindsight-cli/Cargo.toml

@@ -1,13 +1,13 @@
 [package]
-name = "memora-cli-rust"
+name = "hindsight-cli"
 version = "0.0.7"
 edition = "2021"
-authors = ["Memora Team"]
-description = "A beautiful CLI for Memora - semantic memory system"
+authors = ["Hindsight Team"]
+description = "A beautiful CLI for Hindsight - semantic memory system"
 license = "MIT"
 
 [[bin]]
-name = "memora"
+name = "hindsight"
 path = "src/main.rs"
 
 [dependencies]

hindsight-cli/build.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 set -e
 
-# Build script for memora-cli-rust
+# Build script for hindsight-cli
 # This script builds optimized binaries for multiple platforms
 
 # Source cargo environment if it exists
@@ -19,7 +19,7 @@ if ! command -v cargo &> /dev/null; then
     exit 1
 fi
 
-echo "Building Memora CLI for multiple platforms..."
+echo "Building Hindsight CLI for multiple platforms..."
 
 # Ensure we're in the right directory
 cd "$(dirname "$0")"
@@ -50,10 +50,10 @@ build_target() {
 
     # Copy to dist
     if [[ "$target" == *"windows"* ]]; then
-        cp "target/$target/release/memora.exe" "dist/$output_name.exe"
+        cp "target/$target/release/hindsight.exe" "dist/$output_name.exe"
         echo "Created: dist/$output_name.exe"
     else
-        cp "target/$target/release/memora" "dist/$output_name"
+        cp "target/$target/release/hindsight" "dist/$output_name"
         chmod +x "dist/$output_name"
         echo "Created: dist/$output_name"
     fi
@@ -70,19 +70,19 @@ case "$OS" in
     Darwin)
         if [[ "$ARCH" == "arm64" ]]; then
             echo "Building for macOS ARM64 (Apple Silicon)..."
-            build_target "aarch64-apple-darwin" "memora-macos-arm64"
+            build_target "aarch64-apple-darwin" "hindsight-macos-arm64"
         else
             echo "Building for macOS x86_64 (Intel)..."
-            build_target "x86_64-apple-darwin" "memora-macos-x86_64"
+            build_target "x86_64-apple-darwin" "hindsight-macos-x86_64"
         fi
         ;;
     Linux)
         if [[ "$ARCH" == "x86_64" ]]; then
             echo "Building for Linux x86_64..."
-            build_target "x86_64-unknown-linux-gnu" "memora-linux-x86_64"
+            build_target "x86_64-unknown-linux-gnu" "hindsight-linux-x86_64"
         elif [[ "$ARCH" == "aarch64" ]]; then
             echo "Building for Linux ARM64..."
-            build_target "aarch64-unknown-linux-gnu" "memora-linux-arm64"
+            build_target "aarch64-unknown-linux-gnu" "hindsight-linux-arm64"
         fi
         ;;
     *)