pr
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources including pull request descriptions and commit messages. An attacker could potentially embed malicious instructions within these inputs to influence the agent's behavior. * Ingestion points: Pull request title, description, and commit history. * Boundary markers: Not specified. * Capability inventory: Access to pull request metadata and git operations. * Sanitization: None mentioned for external text.
- [COMMAND_EXECUTION]: The skill uses shell commands for git operations with parameters like branch names sourced from pull requests. This is a potential command injection vector if the input is not sanitized. * Evidence: Shell commands in the History Cleanup section.
Audit Metadata