complete-pr
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
ghCLI and Git commands (commit,push) to manage repository workflows and pull request status. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes untrusted input from pull request bodies and review comments.
- Ingestion points: Pull request descriptions and review messages retrieved via the
fetch-prskill. - Boundary markers: None explicitly defined in the instructions to separate untrusted data from instructions.
- Capability inventory: The agent can perform file system writes (code changes), Git commits, Git pushes, and update PR metadata.
- Sanitization: The instructions mitigate risk by directing the agent to validate claims against the codebase and official documentation before acting, though it remains a potential surface for indirect injection.
Audit Metadata