feat: complete phase 5 -- error hardening, frontend, Docker, demo, docs

Backend:
- ConversationTracker: Protocol + PostgresConversationTracker for lifecycle tracking
- Error handler: ErrorCategory enum, classify_error(), with_retry() exponential backoff
- Wire PostgresAnalyticsRecorder + ConversationTracker into ws_handler
- Rate limiting (10 msg/10s per thread), edge case hardening
- Health endpoint GET /api/health, version 0.5.0
- Demo seed data script + sample OpenAPI spec

Frontend (all new):
- React Router with NavBar (Chat / Replay / Dashboard / Review)
- ReplayListPage + ReplayPage with ReplayTimeline component
- DashboardPage with MetricCard, range selector, zero-state
- ReviewPage for OpenAPI classification review
- ErrorBanner for WebSocket disconnect handling
- API client (api.ts) with typed fetch wrappers

Infrastructure:
- Frontend Dockerfile (multi-stage node -> nginx)
- nginx.conf with SPA routing + API/WS proxy
- docker-compose.yml with frontend service + healthchecks
- .env.example files (root + backend)

Documentation:
- README.md with quick start and architecture
- Agent configuration guide
- OpenAPI import guide
- Deployment guide
- Demo script

48 new tests, 449 total passing, 92.87% coverage
This commit is contained in:
Yaojia Wang
2026-03-31 21:20:06 +02:00
parent 38644594d2
commit 0e78e5b06b
44 changed files with 3397 additions and 169 deletions

View File

@@ -5,6 +5,8 @@ from __future__ import annotations
import json
import logging
import re
import time
from collections import defaultdict
from typing import TYPE_CHECKING, Any
from langchain_core.messages import HumanMessage
@@ -16,16 +18,23 @@ if TYPE_CHECKING:
from fastapi import WebSocket
from langgraph.graph.state import CompiledStateGraph
from app.analytics.event_recorder import AnalyticsRecorder
from app.callbacks import TokenUsageCallbackHandler
from app.conversation_tracker import ConversationTrackerProtocol
from app.interrupt_manager import InterruptManager
from app.session_manager import SessionManager
logger = logging.getLogger(__name__)
MAX_MESSAGE_SIZE = 32_768 # 32 KB
MAX_CONTENT_LENGTH = 8_000 # characters
MAX_CONTENT_LENGTH = 10_000 # characters
THREAD_ID_PATTERN = re.compile(r"^[a-zA-Z0-9\-_]{1,128}$")
# Rate limiting: max 10 messages per 10-second window, per thread
_RATE_LIMIT_MAX = 10
_RATE_LIMIT_WINDOW = 10.0
_thread_timestamps: dict[str, list[float]] = defaultdict(list)
async def handle_user_message(
ws: WebSocket,
@@ -197,6 +206,9 @@ async def dispatch_message(
callback_handler: TokenUsageCallbackHandler,
raw_data: str,
interrupt_manager: InterruptManager | None = None,
analytics_recorder: AnalyticsRecorder | None = None,
conversation_tracker: ConversationTrackerProtocol | None = None,
pool: Any = None,
) -> None:
"""Parse and route an incoming WebSocket message."""
if len(raw_data) > MAX_MESSAGE_SIZE:
@@ -205,10 +217,14 @@ async def dispatch_message(
try:
data = json.loads(raw_data)
except json.JSONDecodeError:
except (json.JSONDecodeError, ValueError):
await _send_json(ws, {"type": "error", "message": "Invalid JSON"})
return
if not isinstance(data, dict):
await _send_json(ws, {"type": "error", "message": "Invalid JSON: expected object"})
return
msg_type = data.get("type")
thread_id = data.get("thread_id", "")
@@ -222,16 +238,36 @@ async def dispatch_message(
if msg_type == "message":
content = data.get("content", "")
if not content:
if not content or not content.strip():
await _send_json(ws, {"type": "error", "message": "Missing message content"})
return
if len(content) > MAX_CONTENT_LENGTH:
await _send_json(ws, {"type": "error", "message": "Message content too long"})
return
# Rate limiting check
now = time.time()
timestamps = _thread_timestamps[thread_id]
cutoff = now - _RATE_LIMIT_WINDOW
_thread_timestamps[thread_id] = [t for t in timestamps if t >= cutoff]
if len(_thread_timestamps[thread_id]) >= _RATE_LIMIT_MAX:
await _send_json(ws, {"type": "error", "message": "Rate limit exceeded"})
return
_thread_timestamps[thread_id].append(now)
await handle_user_message(
ws, graph, session_manager, callback_handler, thread_id, content,
interrupt_manager=interrupt_manager,
)
await _fire_and_forget_tracking(
thread_id=thread_id,
pool=pool,
analytics_recorder=analytics_recorder,
conversation_tracker=conversation_tracker,
agent_name=None,
tokens=0,
cost=0.0,
)
elif msg_type == "interrupt_response":
approved = data.get("approved", False)
@@ -244,6 +280,36 @@ async def dispatch_message(
await _send_json(ws, {"type": "error", "message": "Unknown message type"})
async def _fire_and_forget_tracking(
thread_id: str,
pool: Any,
analytics_recorder: Any | None,
conversation_tracker: Any | None,
agent_name: str | None,
tokens: int,
cost: float,
) -> None:
"""Fire-and-forget analytics/tracking; failures must NOT break chat."""
try:
if conversation_tracker is not None and pool is not None:
await conversation_tracker.ensure_conversation(pool, thread_id)
await conversation_tracker.record_turn(pool, thread_id, agent_name, tokens, cost)
except Exception:
logger.exception("Conversation tracker error for thread %s (suppressed)", thread_id)
try:
if analytics_recorder is not None:
await analytics_recorder.record(
thread_id=thread_id,
event_type="message",
agent_name=agent_name,
tokens_used=tokens,
cost_usd=cost,
)
except Exception:
logger.exception("Analytics recorder error for thread %s (suppressed)", thread_id)
def _has_interrupt(state: Any) -> bool:
"""Check if the graph state has a pending interrupt."""
tasks = getattr(state, "tasks", ())