washingtonguilhardes/monitor

MONITOR Task

Persona: Execute this task as the @developer subagent. Load the persona characteristics from .rulesync/subagents/developer.md before proceeding.

Objective

Add structured logging and error tracking to improve observability, debugging, monitoring.

Instructions

1. Discovery: - Ask: Code to monitor? (file/function/feature), Monitoring level? (basic logging/detailed tracing/error tracking), Specific error scenarios?, Error boundaries for frontend?

2. Review existing: - Check console.log/console.error usage - Review error handling in tRPC procedures - Examine TRPCError patterns

3. Add structured logging (tRPC): - Log at entry: console.log("[procedure] Starting", {userId, input}) - Log success: console.log("[procedure] Success", {result}) - Log errors: console.error("[procedure] Failed", {error, context}) - Use TRPCError with appropriate codes

4. Error boundaries (React): - Create ErrorBoundary component: getDerivedStateFromError, componentDidCatch with logging - Use: <ErrorBoundary name="ComponentName" fallback={...}>{children}</ErrorBoundary>

5. TRPCError usage: - NOT_FOUND: Channel/user not found - FORBIDDEN: Not authorized - BAD_REQUEST: Validation error - INTERNAL_SERVER_ERROR: Operation failed (with cause)

6. Logging points: - Procedure entry (with input), Database operations, Business logic decisions, Errors/exceptions, Procedure exit (with results)

7. Security: - Never log: passwords, API tokens - Caution with: user data in logs - Log: error codes, not sensitive details - Use structured logging (JSON objects)

8. Performance monitoring: - Track execution time for slow operations - Log duration, flag if > threshold

9. Test error scenarios: - Add tests to verify error handling - Test error paths ensure handled correctly

10. Quality checks: - pnpm lint && pnpm typecheck && pnpm test

11. Summary: - Procedures/components instrumented - Error boundaries added - Logging added (entry/exit/errors/performance) - Security considerations - Testing coverage - Log analysis examples

Example Prompts

• /monitor app/api/trpc/routers/channels.ts --level detailed - Adds structured entry/exit/error logging to each procedure, tracks execution duration, and uses TRPCError codes consistently.
• /monitor app/ --level basic --error-boundaries - Adds ErrorBoundary components around key UI sections with meaningful fallback UI and error logging to console.

Quality Criteria

• Logging uses structured format with JSON objects (e.g., {procedure, userId, duration}) not string concatenation
• Sensitive data (passwords, tokens, API keys, PII) is never logged -- even at debug level
• Error boundaries include meaningful fallback UI that tells the user what went wrong, not just a generic "Something went wrong"
• Performance thresholds are documented next to timing checks (e.g., // warn if > 500ms)
• TRPCError codes match the semantic meaning: NOT_FOUND, FORBIDDEN, BAD_REQUEST, INTERNAL_SERVER_ERROR
• Every procedure has at minimum entry and error logging after instrumentation

Scope Boundaries

• Add observability (logging, error boundaries, timing) -- do NOT refactor business logic
• Add error boundaries -- do NOT change existing error handling behavior or catch logic
• Do NOT introduce new dependencies or logging libraries without discussion
• Do NOT modify test assertions unless a test explicitly validates logging output

Cross-References

• /audit - Identifies observability gaps and missing error handling
• /review - Checks logging quality and security of log statements