feature-discovery
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from source code and issue trackers. It mitigates potential injection risks by enforcing a rigid, read-only workflow, requiring structured output via a mandatory report template, and mandating that the agent stop after presenting the report to wait for explicit user approval before any further action.
- [COMMAND_EXECUTION]: The workflow involves executing standard CLI tools like
rg(ripgrep) andgh(GitHub CLI) for information retrieval. These are strictly confined to read-only operations within the mapped repository scope, with explicit rules forbidding destructive actions, migrations, or installation commands.
Audit Metadata