leaving-pr-comments
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted data from pull requests, creating a surface for indirect prompt injection.\n
- Ingestion points: The workflow involves reading the pull request head, code diffs, comment history, and thread data (SKILL.md).\n
- Boundary markers: The instructions do not define specific delimiters or directives to ignore instructions embedded within the ingested data.\n
- Capability inventory: The skill possesses permissions to post top-level comments, reply to existing threads, and resolve discussion threads on the repository (SKILL.md).\n
- Sanitization: While the instructions recommend checking for private data, they lack specific measures to sanitize or escape pull request content to mitigate instructional injection.
Audit Metadata