acp-conformance
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill fetches protocol specifications and production readiness guides from trusted domains including OpenAI's official developer portal and GitHub repositories.
- [SAFE]: No malicious command execution, remote script downloads, or persistence mechanisms were detected in the instructions.
- [SAFE]: The skill involves the ingestion of external data via web search and fetching for documentation purposes. This ingestion surface is considered safe as it targets official sources. Ingestion points: SKILL.md (fetching guides and schemas); Boundary markers: None; Capability inventory: Bash, Write, Edit, Read; Sanitization: None.
- [SAFE]: Protocol testing instructions regarding headers (e.g., Authorization: Bearer ) are descriptive requirements for testing an implementation and do not contain hardcoded credentials.
Audit Metadata