receiving-code-review
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest code review feedback from external sources and implement the suggested changes, creating a vulnerability to malicious instructions embedded in the feedback.
- Ingestion Points: Feedback from 'External Reviewers' described in
SKILL.mdis processed as instructions for implementation. - Boundary Markers: The skill employs behavioral boundary markers, such as instructions to 'Verify before implementing' and 'be skeptical,' which act as a safeguard against immediate obedience to external input.
- Capability Inventory: The skill involves capabilities for reading the codebase (
grep), implementing logic changes, and interacting with the GitHub API (gh api). - Sanitization: There is no mention of automated sanitization or filtering of the feedback text before the agent processes it.
Audit Metadata