Finds source-visible semantic runtime bugs without running tests or builds. Use for repository bug hunts from an agent, Codex automation, or Claude loop that should report actionable correctness, security, dataflow, state/lifecycle, API-contract, validation, cache/persistence, or error-path bugs, not style, lint, architecture, maintainability, or missing-test feedback.
74
93%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Passed
No findings from the security scan
DEVANA-FINDING: v1 DEVANA-STATE: open | P2 | medium | security=no DEVANA-KEY: path/to/file.ext:123 | short-slug
Describe the concrete issue in plain language.
Name the expected behavior that appears to be broken.
Name the evidence source: docs, schema, tests, caller/callee contract, neighboring implementation, or implicit safety expectation.
Give the smallest concrete value, state, event sequence, or failure path that could trigger the bug.
Explain likely impact without overstating certainty.
List the dataflow trace, control-flow trace, counterexample value, contract mismatch, cross-entry mismatch, or state transition mismatch that makes this actionable.
List guards, callers, config, framework behavior, transactions, cleanup, and the strongest reason this might be false. Explain why that counterevidence does not prevent the bug.
Describe the smallest useful human follow-up.
After working this report, preserve the original finding body. Update line 2 DEVANA-STATE: ... and the final DEVANA-SUMMARY: status/priority/confidence prefix. Use one of: open, fixed, invalid, stale, duplicate, wontfix. Keep DEVANA-KEY: stable unless the same finding moved. Add dated notes below with evidence checked.
DEVANA-KEY: path/to/file.ext:123 | short-slug DEVANA-SUMMARY: open | P2 | medium | One sentence explaining the bug and likely impact.