What problem does it solve? When your worktree's isolated Docker stack misbehaves — the API crashes, a Prisma migration doesn't apply, or an env change never takes effect — you need a fast, structured way to inspect containers, logs, and the auto-apply bridge without fighting timeouts or stale container names. ## Core Features & Use Cases - One-shot stack snapshot: Run pnpm stack:debug to get a JSON status of project, postgres, api, web, and bridge services. - Log inspection without truncation: Use pnpm stack:logs and pnpm stack:follow for full or live service logs, and pnpm stack:bridge to read the NDJSON bridge event log covering migrations, env applies, and rebuilds. - Container exec and recovery: Exec into containers to verify env vars, and follow the decision table to fix a dead bridge via the idempotent stack-up-docker.sh respawn. - Use Case: Your pnpm probe:smoke run reports 500s. You run pnpm stack:logs api | grep -i "error", then pnpm stack:follow api to reproduce the failing request live and pinpoint the crash. ## Quick Start Ask the agent to debug why the API container in your worktree stack is crashing using the stack-debug helpers.
