bug-triage
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill correctly identifies bug reports as untrusted data and provides clear instructions to the agent to disregard any instructions found within them (e.g., 'ignore your task', 'delete tests').
- [DATA_EXPOSURE]: Includes a specific prohibition against accessing or modifying sensitive files like
.env*, which is a best practice for preventing credential harvesting during automated tasks. - [COMMAND_EXECUTION]: Limits the scope of execution to reading code, writing new regression tests, running existing test suites, and standard git operations (commit/branch/PR). It explicitly forbids dangerous actions like merging or pushing to default branches.
- [METADATA_POISONING]: The YAML frontmatter accurately describes the skill's purpose and scope without deceptive instructions.
Audit Metadata