Initial commit: standalone vector memory server

REST API for per-agent semantic memory retrieval.

- FastAPI server with /api/{agent}/save and /api/{agent}/query
- ChromaDB for persistent vector storage
- all-MiniLM-L6-v2 via sentence-transformers for embeddings
- Per-agent collections for clean separation
- Config through env vars or config.py
- .venv ready with all dependencies

main.py

@@ -0,0 +1,109 @@
+"""Vector Memory Server — REST API.
+
+Endpoints
+---------
+- ``POST /api/{agent_name}/save``   — Save a memory
+- ``POST /api/{agent_name}/query``  — Query memories by semantic similarity
+- ``GET  /health``                   — Health check
+"""
+
+from __future__ import annotations
+
+from typing import Optional
+
+import uvicorn
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+
+from config import Config
+from store import MemoryStore
+
+# ---------------------------------------------------------------------------
+# App
+# ---------------------------------------------------------------------------
+
+app = FastAPI(
+    title="Vector Memory Server",
+    description="Semantic memory retrieval via REST. "
+    "Per-agent collections, powered by ChromaDB + all-MiniLM-L6-v2.",
+    version="0.1.0",
+)
+
+store = MemoryStore(Config.DATA_DIR)
+
+
+# ---------------------------------------------------------------------------
+# Request / Response models
+# ---------------------------------------------------------------------------
+
+class SaveRequest(BaseModel):
+    text: str
+    type: Optional[str] = "fact"
+    tags: Optional[list[str]] = None
+    date: Optional[str] = None
+
+
+class SaveResponse(BaseModel):
+    success: bool
+    id: str
+
+
+class QueryRequest(BaseModel):
+    query: str
+    top_n: Optional[int] = Config.DEFAULT_TOP_N
+
+
+class QueryResponse(BaseModel):
+    results: list[dict]
+
+
+# ---------------------------------------------------------------------------
+# Routes
+# ---------------------------------------------------------------------------
+
+@app.post("/api/{agent_name}/save", response_model=SaveResponse)
+async def save_memory(agent_name: str, request: SaveRequest) -> SaveResponse:
+    """Embed and store a memory for the given agent."""
+    try:
+        metadata: dict = {}
+        if request.type:
+            metadata["type"] = request.type
+        if request.tags:
+            metadata["tags"] = ",".join(request.tags)
+        if request.date:
+            metadata["date"] = request.date
+
+        memory_id = store.save(agent_name, request.text, metadata)
+        return SaveResponse(success=True, id=memory_id)
+    except Exception as exc:
+        raise HTTPException(status_code=500, detail=str(exc))
+
+
+@app.post("/api/{agent_name}/query", response_model=QueryResponse)
+async def query_memory(
+    agent_name: str, request: QueryRequest
+) -> QueryResponse:
+    """Retrieve semantically similar memories for the given agent."""
+    try:
+        results = store.query(agent_name, request.query, request.top_n)
+        return QueryResponse(results=results)
+    except Exception as exc:
+        raise HTTPException(status_code=500, detail=str(exc))
+
+
+@app.get("/health")
+async def health() -> dict:
+    return {"status": "ok"}
+
+
+# ---------------------------------------------------------------------------
+# Entrypoint
+# ---------------------------------------------------------------------------
+
+if __name__ == "__main__":
+    uvicorn.run(
+        "main:app",
+        host=Config.HOST,
+        port=Config.PORT,
+        reload=False,
+    )