harness-engineering

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill audits and processes content from the user's project to generate new instructions and configurations. \n
  • Ingestion points: The Audit phase scans existing project files, including CLAUDE.md, README.md, and the docs/ directory. \n
  • Boundary markers: No explicit instructions are provided to delimit external data from agent instructions during the audit or generation phases. \n
  • Capability inventory: The skill generates new documentation files, architectural maps, and linter configurations. \n
  • Sanitization: The skill does not explicitly specify sanitization of codebase content before including it in generated artifacts. \n- [DYNAMIC_EXECUTION]: The skill creates executable configurations and scripts to enforce architectural constraints. \n
  • Evidence: The skill provides templates for GitHub Actions (.github/workflows/code-gc.yml) and custom TypeScript linters (.linters/layer-dependencies.ts). \n
  • Context: This behavior is consistent with the skill's primary purpose of setting up mechanical enforcement and CI-based quality checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 11:18 AM
Security Audit — agent-trust-hub — harness-engineering