pr-review
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a pull request review workflow for the 'saadeghi/daisyui' repository. It operates under a strict read-only policy for GitHub actions and restricts file writing to a dedicated local directory ('tmp/pr/'), preventing unauthorized modifications to the repository or product code.
- [COMMAND_EXECUTION]: The skill instructions allow for the execution of code from pull requests to verify fixes and regressions. This behavior is mitigated by explicit hard rules requiring the agent to manually inspect code before execution and to use isolated, temporary OS environments that lack access to credentials, environment variables, or sensitive system files.
- [PROMPT_INJECTION]: The skill ingests untrusted data from pull request descriptions, comments, and code changes, which represents a potential surface for indirect prompt injection.
- Ingestion points: Pull request diffs, issue descriptions, and user comments from the 'saadeghi/daisyui' repository (SKILL.md).
- Boundary markers: The skill mandates manual code inspection and the use of clean, isolated source copies for testing (SKILL.md).
- Capability inventory: The agent can execute shell commands to run package tests, builds, and reproductions (SKILL.md).
- Sanitization: No specific input sanitization is defined for processing PR content, but the agent is explicitly instructed to ignore author claims and rely solely on verified code behavior and observed command output.
Audit Metadata