Problem or Motivation
Current deploy verification only checks if the container status is "running" — it doesn't verify the app inside is actually healthy (bot connected to Telegram, MCP servers initialized, DB migrated). A container can be "running" but the app inside crashed and is in a restart loop.
Proposed Solution
-
Health check endpoint — add a simple HTTP health server (e.g., on port 8080) that reports:
- Bot connected to Telegram (polling active)
- MCP manager initialized
- DB accessible
- Last successful message timestamp
-
Docker HEALTHCHECK — add to Dockerfile:
HEALTHCHECK --interval=10s --timeout=3s --retries=3 \
CMD curl -f http://localhost:8080/health || exit 1
-
Deploy verification upgrade — check health endpoint instead of just container status
-
Version endpoint — expose current version from package.json at /version for deployment tracking
Acceptance Criteria
Problem or Motivation
Current deploy verification only checks if the container status is "running" — it doesn't verify the app inside is actually healthy (bot connected to Telegram, MCP servers initialized, DB migrated). A container can be "running" but the app inside crashed and is in a restart loop.
Proposed Solution
Health check endpoint — add a simple HTTP health server (e.g., on port 8080) that reports:
Docker HEALTHCHECK — add to Dockerfile:
Deploy verification upgrade — check health endpoint instead of just container status
Version endpoint — expose current version from package.json at
/versionfor deployment trackingAcceptance Criteria
/healthreturns JSON with component status/versionreturns current version and git SHA