artifact-production
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill identifies and mitigates risks associated with processing untrusted document templates, source data, and spreadsheet content. Evidence of ingestion points exists in SKILL.md (Establish the deliverable contract) and references/format-verification.md (Spreadsheet). The skill provides explicit boundary markers and sanitization guidance, instructing the agent to treat embedded macros, scripts, and formula introducers as untrusted input. No exploitable tool capabilities or dangerous subprocess calls are defined in the provided files.
- [EXTERNAL_DOWNLOADS]: The documentation in references/format-verification.md references the official OpenAI GitHub repository for its PDF verification guidance. This is documented as a neutral reference to a trusted organization and does not include automated download or execution patterns.
Audit Metadata