pre-execution-review
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFENO_CODEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill acts as an internal reference for pre-execution review policies and ledger structures. It defines organizational rules and technical standards rather than providing executable tools or agent commands.
- [NO_CODE]: The skill contains markdown policy documentation and configuration schemas, but no executable code within the provided skill files.
- [EXTERNAL_DOWNLOADS]: The skill references the @gtrabanco/agentic-workflow-schema NPM package, which is a vendor-owned resource used for formalizing pre-execution artifact snapshots and ensuring schema consistency.
- [INDIRECT_PROMPT_INJECTION]: The documentation identifies the risk of untrusted content within reviewed artifacts like SPEC.md. It implements a specific policy (§7) that mandates treating all artifact content as data rather than instructions, requiring agents to report embedded directives as defects. This establishes clear ingestion points (unit artifacts), capability constraints, boundary markers (policy §7), and sanitization logic (distinguishing data from commands).
Audit Metadata