Add AI Patronus browser memory example

examples/README.md

@@ -21,6 +21,7 @@ service keys.
 | [A Stripe Link checkout with an SIE fraud-risk gate](./stripe-link-fraud) | Wiring all three SIE primitives into a pre-authorization fraud-risk gate that runs in the same round-trip as the Stripe PaymentIntent | `extract`, `encode`, `score` | Docker Compose plus Node UI; Stripe test-mode keys optional (runs in mock mode without them) | Runnable demo |
 | [Vision-first document RAG](./vision-doc-rag) | Retrieving and answering questions over a multi-tenant page corpus by looking at page images — including scanned drawings — with OCR kept out of the score path | `encode`, `chat/completions`, `score` (optional) | GPU SIE deployment required: ColQwen2.5 retriever + Qwen3.5-4B answer model (runs on the generation bundle) | Runnable demo |
 | [Multi-model contract review with the OpenAI Agents SDK](./contract-review-agent) | Running an OpenAI Agents SDK agent whose every model call — triage, orchestration, vision, OCR, embeddings, rerank, entity extraction, text-to-SQL, reasoning, and a safety guardrail — is served by one SIE cluster, each step on the right catalog model, with per-model observability | `chat/completions`, `encode`, `score`, `extract` | GPU SIE deployment required; standalone `uv` project; real contracts fetched from CUAD (CC BY 4.0) | Runnable demo |
+| [AI Patronus browser memory](./ai-patronus-browser-memory) | Building a browser companion that saves pages on request and recalls them semantically | OpenAI-compatible `/v1/embeddings` | Local SIE Docker image plus Chrome extension and FastAPI bridge | Runnable demo |
 
 For docs publishing, lead with the quickest runnable demos, then use the
 benchmark and evaluation examples for deeper technical users.

examples/ai-patronus-browser-memory/.gitignore

@@ -0,0 +1,7 @@
+.env
+.venv/
+__pycache__/
+memory_server/data/
+memory_server/.env
+memory_server/.venv/
+extension/config.local.js

examples/ai-patronus-browser-memory/README.md

@@ -0,0 +1,147 @@
+# AI Patronus browser memory
+
+AI Patronus is a Chrome extension that lets a small browser companion remember
+pages on request and recall them semantically through SIE embeddings.
+
+This example focuses on a privacy-preserving memory loop:
+
+1. The user asks Patronus to `remember this page`.
+2. The extension sends the current page title, URL, and readable text to a local
+   memory bridge.
+3. The bridge calls SIE's OpenAI-compatible `/v1/embeddings` endpoint and stores
+   the vector locally.
+4. The user asks `what did I save about browser memory?` and Patronus returns the
+   closest saved pages.
+
+No browser history is uploaded automatically. The save action is manual, and the
+default setup runs against a local SIE server with local JSON storage.
+
+## SIE primitives
+
+| Flow | Endpoint | Model |
+|---|---|---|
+| Page memory | `/v1/embeddings` | `sentence-transformers/all-MiniLM-L6-v2` |
+| Semantic recall | `/v1/embeddings` + local cosine similarity | `sentence-transformers/all-MiniLM-L6-v2` |
+
+The Chrome extension also contains optional hooks for chat, voice, and web
+research providers from the original hackathon prototype, but the SIE memory
+flow works without those keys.
+
+## Run it locally
+
+You need Docker, Python 3.12, and Chrome.
+
+Start SIE:
+
+```bash
+docker run -p 8080:8080 -v sie-hf-cache:/app/.cache/huggingface ghcr.io/superlinked/sie-server:latest-cpu-default
+```
+
+Start the memory bridge:
+
+```bash
+cd examples/ai-patronus-browser-memory/memory_server
+python -m venv .venv
+. .venv/bin/activate
+pip install -r requirements.txt
+cp .env.example .env
+uvicorn app:app --reload --port 8800
+```
+
+Load the extension:
+
+1. Open `chrome://extensions`.
+2. Enable Developer mode.
+3. Choose **Load unpacked**.
+4. Select `examples/ai-patronus-browser-memory/extension`.
+5. Open the Patronus popup and set `Superlinked URL` to `http://localhost:8800`.
+
+Try it on any normal website:
+
+```text
+remember this page
+what did I save about browser memory?
+```
+
+You can also seed sample memories after the memory bridge is running:
+
+```bash
+cd examples/ai-patronus-browser-memory/memory_server
+python seed_sample.py
+```
+
+Then ask:
+
+```text
+what did I save about privacy?
+```
+
+## Configuration
+
+The memory bridge reads `memory_server/.env`:
+
+| Variable | Default | Purpose |
+|---|---|---|
+| `SIE_URL` | `http://localhost:8080` | SIE endpoint |
+| `SIE_API_KEY` | empty | Optional bearer token for auth-enabled SIE clusters |
+| `SIE_EMBED_MODEL` | `sentence-transformers/all-MiniLM-L6-v2` | Embedding model |
+| `PATRONUS_MEMORY_PATH` | `data/memory.json` | Local memory store |
+| `PATRONUS_MAX_TEXT_CHARS` | `12000` | Per-page text cap |
+
+The extension popup can store optional provider keys in Chrome local storage.
+For a hosted/auth-enabled SIE memory endpoint, add that endpoint's origin to
+`extension/manifest.json` `host_permissions` before loading the extension.
+
+## API
+
+The local bridge exposes:
+
+### `POST /ingest`
+
+```json
+{
+  "title": "Example page",
+  "url": "https://example.com",
+  "text": "Page text to remember"
+}
+```
+
+### `POST /query`
+
+```json
+{
+  "query": "browser memory",
+  "limit": 5
+}
+```
+
+Response:
+
+```json
+{
+  "results": [
+    {
+      "title": "Example page",
+      "url": "https://example.com",
+      "text": "Short excerpt...",
+      "score": 0.82
+    }
+  ]
+}
+```
+
+## Project layout
+
+```text
+ai-patronus-browser-memory/
+├── extension/          # Manifest V3 Chrome extension
+└── memory_server/     # FastAPI bridge from extension to SIE embeddings
+```
+
+## Privacy notes
+
+- The extension sends page text only when the user explicitly asks it to
+  remember the page.
+- The bridge stores memories locally by default in `memory_server/data/`.
+- Do not enable automatic history ingestion without explicit consent, filtering,
+  and deletion controls.