orch-pipeline
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to its ingestion of external data.\n- Ingestion points: Phase 0 (intake of external design documents) and Phase 1 (GitHub search results and external registry data).\n- Boundary markers: Not explicitly defined in the pipeline metadata, though human review is required before execution.\n- Capability inventory: File system modification, code execution (TDD), and git commit operations.\n- Sanitization: The skill relies on human oversight rather than automated sanitization to validate ingested content.
Audit Metadata