Initial commit: Multi-service AI agent system

- Frontend: Vite + React + TypeScript chat interface
- Backend: FastAPI gateway with LangGraph routing
- Knowledge Service: ChromaDB RAG with Gitea scraper
- LangGraph Service: Multi-agent orchestration
- Airflow: Scheduled Gitea ingestion DAG
- Documentation: Complete plan and implementation guides

Architecture:
- Modular Docker Compose per service
- External ai-mesh network for communication
- Fast rebuilds with /app/packages pattern
- Intelligent agent routing (no hardcoded keywords)

Services:
- Frontend (5173): React chat UI
- Chat Gateway (8000): FastAPI entry point
- LangGraph (8090): Agent orchestration
- Knowledge (8080): ChromaDB RAG
- Airflow (8081): Scheduled ingestion
- PostgreSQL (5432): Chat history

Excludes: node_modules, .venv, chroma_db, logs, .env files
Includes: All source code, configs, docs, docker files

backend/tests/test_gateway.py

@@ -0,0 +1,79 @@
+import pytest
+from fastapi.testclient import TestClient
+from main import app
+import httpx
+from unittest.mock import AsyncMock, patch
+
+client = TestClient(app)
+
+@pytest.mark.asyncio
+async def test_chat_general_query():
+    """Test that a general query (no personal keywords) skips the Librarian."""
+    with patch("httpx.AsyncClient.post", new_callable=AsyncMock) as mock_post:
+        # Mock Brain response
+        mock_response = AsyncMock()
+        mock_response.status_code = 200
+        mock_response.json.return_value = {
+            "info": {"id": "msg_123"},
+            "parts": [{"type": "text", "text": "I am a general AI."}]
+        }
+        
+        # First call is for session creation, second for message
+        mock_post.side_effect = [AsyncMock(status_code=200, json=lambda: {"id": "ses_123"}), mock_response]
+        
+        response = client.post("/chat", json={"message": "What is 2+2?"})
+        
+        assert response.status_code == 200
+        assert response.json()["response"] == "I am a general AI."
+        # Verify Librarian (knowledge-service) was NOT called
+        # The knowledge service URL is http://knowledge-service:8080/query
+        calls = [call.args[0] for call in mock_post.call_args_list]
+        assert not any("knowledge-service" in url for url in calls)
+
+@pytest.mark.asyncio
+async def test_chat_personal_query_success():
+    """Test that a personal query calls the Librarian and injects context."""
+    with patch("httpx.AsyncClient.post", new_callable=AsyncMock) as mock_post:
+        # 1. Mock Librarian Response
+        mock_k_res = AsyncMock()
+        mock_k_res.status_code = 200
+        mock_k_res.json.return_value = {"context": "Sam likes red guitars."}
+        
+        # 2. Mock Brain Session Response
+        mock_s_res = AsyncMock()
+        mock_s_res.status_code = 200
+        mock_s_res.json.return_value = {"id": "ses_123"}
+        
+        # 3. Mock Brain Message Response
+        mock_b_res = AsyncMock()
+        mock_b_res.status_code = 200
+        mock_b_res.json.return_value = {
+            "parts": [{"type": "text", "text": "I see Sam likes red guitars."}]
+        }
+        
+        mock_post.side_effect = [mock_k_res, mock_s_res, mock_b_res]
+        
+        response = client.post("/chat", json={"message": "Tell me about Sam's music"})
+        
+        assert response.status_code == 200
+        assert "red guitars" in response.json()["response"]
+        
+        # Verify Librarian was called
+        calls = [call.args[0] for call in mock_post.call_args_list]
+        assert any("knowledge-service" in url for url in calls)
+
+@pytest.mark.asyncio
+async def test_chat_librarian_timeout_failover():
+    """Test that the gateway fails over instantly (5s) if Librarian is slow."""
+    with patch("httpx.AsyncClient.post", new_callable=AsyncMock) as mock_post:
+        # Mock Librarian Timeout
+        mock_post.side_effect = [
+            httpx.TimeoutException("Timeout"), # Librarian call
+            AsyncMock(status_code=200, json=lambda: {"id": "ses_123"}), # Brain Session
+            AsyncMock(status_code=200, json=lambda: {"parts": [{"type": "text", "text": "Direct Brain Response"}]}) # Brain Msg
+        ]
+        
+        response = client.post("/chat", json={"message": "Sam's hobbies?"})
+        
+        assert response.status_code == 200
+        assert response.json()["response"] == "Direct Brain Response"