pr-review-html
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
gitandgh(GitHub CLI) to retrieve code diffs and pull request information. It also authorizes the agent to run 'relevant validation commands' based on the repository context, which involves executing local build or test scripts as part of the review process. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of pull request diffs and patches, creating a surface for indirect instructions to influence agent behavior during the review or validation phase.
- Ingestion points: Retrieves external data via
gh pr diff,git diff, or provided patch files (SKILL.md). - Boundary markers: The instructions do not specify any markers or delimiters to help the agent distinguish between code and malicious natural language instructions within the diff content.
- Capability inventory: Executes shell commands (
git,gh, validation tools) and performs file system writes to generate the HTML artifact (SKILL.md). - Sanitization: There are no explicit requirements for sanitizing or escaping the content of the diffs before they are rendered into the final HTML report or processed during the analysis.
Audit Metadata