error-handling

Files (2)

• SKILL.md
• SKILL.md.lock

SKILL.md (first lines)

<!-- bodyRef: skills/sources/error-handling.md -->

<!-- Generated by `harness skills bootstrap`. Edit to tune; re-run to regenerate. -->
<!-- Static skills point at live docs by design — do not hardwire version-pinned API here. -->

Use this skill when designing error handling, logging, or failure recovery.

## Concepts (timeless)
- Fail fast at boundaries with clear messages; don't let bad state propagate.
- Distinguish expected errors (validation, not-found) from unexpected ones (bugs, outages); handle each differently.
- Never silently swallow errors. Either handle, wrap-with-context, or propagate.
- User-facing messages are safe and actionable; detailed diagnostics go to structured logs server-side.
- Make recovery explicit: timeouts, retries with backoff + jitter, and a fallback/degraded path.

## Pull current docs first
Error/exception idioms and logging libraries are language- and framework-specific:
- your language's current error/exception + logging library docs

## Checklist
- [ ] Inputs validated; expected vs unexpected errors handled distinctly
- [ ] Errors wrapped with context as they cross layers (no bare re-throw)
- [ ] No secrets/PII in error messages or logs
- [ ] Retries bounded with backoff; timeouts on every I/O call
- [ ] A tested failure/degraded path exists for critical flows

## Gotchas

_Highest-signal section. Seed empty; grow it from real failures this skill
hits — `harness skills propose` suggests entries here from drift evidence._

- _none observed yet_

From ichiorca/shipsignal, Apache-2.0.

More from ichiorca/shipsignal