review-componentization

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute git rev-parse --show-toplevel to programmatically identify the repository root directory.
  • Evidence: Found in Section 1, 'Assemble the judge's inputs'.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted artifacts (HTML, code files) through a sub-agent without explicit security boundaries.
  • Ingestion points: Section 1 specifies gathering 'settled study HTML', 'inventory', and 'built code' files.
  • Boundary markers: The instructions do not define delimiters or specific 'ignore instructions' warnings for the sub-agent's prompt.
  • Capability inventory: The skill uses the Agent tool to spawn sub-tasks and supports modifying code via the 'Fix' directive.
  • Sanitization: No sanitization of ingested content is mentioned before it is passed to the sub-agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 09:20 AM
Security Audit — agent-trust-hub — review-componentization