workflow-end
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes project-local scripts and git commands to audit workflow state and manage baseline records. Evidence: Calls to
.claude/hooks/lib/review-receipt.cjs,.claude/scripts/code_graph,.claude/scripts/lib/workflow-baseline.cjs, andgit diff. These are internal tools used for workflow orchestration.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources to verify mandatory quality gates. \n - Ingestion points: Reads from
plan.md(Quality Gates checklist) and various report files undertmp/reports/(e.g.,review.md). \n - Boundary markers: No explicit delimiters are used to isolate untrusted data from instructions. \n
- Capability inventory: Access to shell command execution through Node.js and Python environments, as well as file system access and git operations. \n
- Sanitization: Verification logic relies on matching status fields (e.g.,
APPROVED) but does not explicitly describe sanitization of the broader file contents before processing.
Audit Metadata