rajudandigam/incident-response-coordinator

# Incident Response Coordinator A TypeScript implementation of the multi-agent incident response system from `.cursor/development_guide.md`. It coordinates diagnostic, database, network, and scaling agents, records local execution traces with `agent-inspect`, and validates behavior with unit, integration, and end-to-end tests. ## Setup npm install cp .env.example .env The default `LLM_MODE=fake` runs deterministic local responses and does not require an API key. Set `LLM_MODE=openai` and `OPENAI_API_KEY` to run against OpenAI. Redis is optional for automated tests. Runtime demos use an in-memory lock manager by default; set `LOCK_MODE=redis` and `REDIS_URL` to exercise Redis-backed locks. ## Run npm run dev -- --scenario conflicting-critical Available scenarios: - `high-latency` - `network-degraded` - `scaling-pressure` - `conflicting-critical` - `healthy` Traces are written to `.agent-inspect/` by default. npx agent-inspect list --dir ./.agent-inspect npx agent-inspect view --dir ./.agent-inspect ## Validate npm run format:check npm run lint npm run typecheck npm test npm run test:e2e npm run build Run the optional Redis smoke test by starting Redis and setting `REDIS_URL`: docker run -p 6379:6379 redis:alpine LOCK_MODE=redis REDIS_URL=redis://localhost:6379 npm run test:e2e

标签：自动化攻击