English | 中文

Hybrid intelligent event trigger system that combines a traditional rules engine with LLM-based reasoning to monitor events and send notifications.

• FastAPI API for rule management, testing, and history.
• Hybrid rule engine: expression-based filters + LLM validation.
• RabbitMQ for event ingestion; Redis for state, context, and throttling.
• Notification channels for email, Telegram, and WeCom (extensible).
• Observability hooks for metrics and tracing.

• Python 3.12+
• Docker (for Redis + RabbitMQ)
• uv (recommended) or another Python environment manager

cp .env.example .env
docker-compose up -d
uv sync --dev

uv run uvicorn llmtrigger.api.app:app --reload --port 8203

API docs: http://localhost:8203/docs

Settings are loaded from environment variables and .env. Common options:

• REDIS_URL, RABBITMQ_URL, RABBITMQ_QUEUE
• OPENAI_API_KEY, OPENAI_BASE_URL, OPENAI_MODEL
• NOTIFICATION_MAX_RETRY, NOTIFICATION_DEFAULT_COOLDOWN

Keep secrets out of version control and use .env locally.

• llmtrigger/api/: FastAPI app, dependencies, and routes
• llmtrigger/engine/: rule engine and LLM logic
• llmtrigger/messaging/: RabbitMQ consumer and handlers
• llmtrigger/storage/: Redis clients and stores
• llmtrigger/notification/: notification dispatcher, worker, channels
• llmtrigger/core/: config and logging
• tests/: unit and integration tests
• docs/: architecture and design notes

uv run pytest

Tests use pytest + pytest-asyncio and live under tests/.

See AGENTS.md for contributor guidelines, style, and PR expectations.

MIT License. See LICENSE.