finding-discovery

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • Command Execution: The skill utilizes bundled Python scripts to assist in the analysis workflow.
  • Evidence: SKILL.md specifies the use of <python_command> <plugin_dir>/scripts/generate_rank_input.py for tasks such as generating rank inputs and copying review data.
  • Indirect Prompt Injection Surface: The tool processes untrusted repository content which is a standard part of security auditing.
  • Ingestion points: Source files and commit messages defined in SKILL.md.
  • Boundary markers: Workflows are governed by referenced policy files like scan-artifacts-and-ledger.md.
  • Capability inventory: Local file reading and script execution using environment placeholders.
  • Sanitization: Instructions advise the agent to verify the actual code logic rather than relying on potentially misleading commit narratives.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 04:23 AM
Security Audit — agent-trust-hub — finding-discovery