summarizing-pr-diffs
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing untrusted external data from pull request bodies and diff files.
- Ingestion points: Untrusted data is ingested when the agent reads PR metadata, bodies, and diffs from files (SKILL.md).
- Boundary markers: The instructions do not define specific delimiters or guidelines to treat the ingested data as non-executable text or to ignore embedded instructions.
- Capability inventory: The skill does not define dangerous capabilities such as network access, arbitrary code execution, or sensitive file system modifications beyond reading the specified PR files.
- Sanitization: No input validation or sanitization rules are specified to filter malicious content within the PR data.
Audit Metadata