pr-reviewer
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data in the form of pull request diffs, descriptions, and existing review comments from GitHub or Azure DevOps.
- Ingestion points: Data is ingested from remote PR URLs or local branch contents via the specialist code-review agent invocation described in
SKILL.md. - Boundary markers: The instructions include specific directives for the sub-agent to validate concerns independently and strictly prohibits unauthorized actions like fetching, switching branches, or posting comments without user approval.
- Capability inventory: The skill possesses the capability to modify the local filesystem (
Apply locally) and interact with remote APIs (Post this exact comment) via the agent's tools. - Sanitization: Significant mitigation is present; the skill requires the user to review and confirm every specific code change or remote comment in an editable field before execution, preventing silent automated attacks.
Audit Metadata