answer-pr-reviews
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands for project verification, such as tests, linters, and type-checkers, as defined in the repository's documentation (AGENTS.md) or manifest files.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources that can be controlled by third parties (GitHub review comments and pull request descriptions).
- Ingestion points: Data is fetched via GitHub GraphQL API, specifically
reviewThreads,reviews, andcommentsin Stage 2 ofSKILL.md. - Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded malicious prompts within the fetched comments before passing them to the sub-skill
receiving-code-review. - Capability inventory: The skill has the ability to modify local code, commit changes, push to remote branches, and execute arbitrary shell commands defined in project manifests.
- Sanitization: No specific sanitization or filtering of the content retrieved from GitHub is mentioned before it influences the agent's actions.
Audit Metadata