autofix

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external PR comments and repository-specific instruction files which serve as potential vectors for malicious instructions.\n
  • Ingestion points: Data is ingested from local AGENTS.md files (Step 0) and via GitHub API calls for PR review threads (Step 3).\n
  • Boundary markers: The skill does not use specific delimiters for external content but relies on explicit negative constraints and safety notes in its instructions to the agent.\n
  • Capability inventory: The agent has permissions to edit repository files and execute version control commands via git and the GitHub CLI.\n
  • Sanitization: The skill provides a comprehensive set of sanitization rules in Step 6 to redact sensitive information and ignore command execution prompts from reviewers.\n- [COMMAND_EXECUTION]: The skill performs shell operations using git and the GitHub CLI (gh) to manage the local and remote repository state.\n
  • Evidence: The workflow includes multiple shell command blocks for managing PRs and repository state, such as gh pr list, gh api graphql, and git commit.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 02:56 PM
Security Audit — agent-trust-hub — autofix