plan-docs-sync

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local shell commands (e.g., npm run <cli> -- --help) to extract CLI configuration details. This is performed to verify the accuracy of onboarding documentation and includes a 'when safe' directive to ensure the agent does not execute potentially harmful commands.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes external documentation and code comment data during its audit phase.
  • Ingestion points: Documentation markdown files, environment variable templates (.env.example), and source code files as defined in the phase flow.
  • Boundary markers: The instructions include a 'Preservation Contract' that mandates verifying all information against the code as the source of truth and explicitly forbids fabricating behavior.
  • Capability inventory: The skill is restricted to 'Audit & plan only' and does not possess capabilities for file modification, software installation, or network access in its current configuration.
  • Sanitization: No explicit content sanitization or escaping of the ingested documentation content is described.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 11:19 AM
Security Audit — agent-trust-hub — plan-docs-sync