pull-request

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Local Command Execution: The skill identifies and executes commands from the repository, such as package scripts and workflow steps, to perform 'CI-confidence' checks. This allows the tool to validate changes locally before submission but relies on the integrity of the repository's configuration files. \n- Indirect Prompt Injection Surface: The skill processes untrusted data to generate PR descriptions, presenting a surface for indirect prompt injection. \n
  • Ingestion points: The skill reads git diffs, commit logs, and repository templates to gather context for the pull request. \n
  • Boundary markers: Instructions include a 'public-output guard' to prevent the inclusion of sensitive values, though specific delimiters for the diff content are not detailed. \n
  • Capability inventory: The agent can execute local package scripts, perform git commits, and push branches to remote repositories. \n
  • Sanitization: The workflow requires human approval before external writes and uses predefined templates to structure the output.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 06:27 PM
Security Audit — agent-trust-hub — pull-request