tao-artifacts
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides the structural contract for artifacts that process external data throughout the TAO pipeline.\n
- Ingestion points: The schemas in
references/spec_bundle.schema.jsonandreferences/job_record.schema.jsondefine the ingress points for user-provided job specifications and experiment configurations.\n - Boundary markers: Strict schema validation is enforced using
additionalProperties: falseto reject unexpected data, and dotted key structures are structurally forbidden to prevent common configuration errors.\n - Capability inventory: The skill utilizes standard validation tools like
jsonschemafor artifact verification and maintains job records via localized scripts.\n - Sanitization: The schemas include regex-based validation for all identifier fields and mandate SHA256 integrity checks for all supporting orchestration files.
Audit Metadata