gpt-orchestration-build
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon untrusted external data sources, specifically "audit, issue set, review, or failing-test reports," to drive automated code implementation.
- Ingestion points: The skill reads "supplied findings" and "repository instructions" as the primary input for its execution workflow in SKILL.md.
- Boundary markers: The instructions lack explicit boundary markers or directives to the agent to disregard instructions potentially embedded within the findings or audit reports.
- Capability inventory: The skill possesses significant capabilities, including modifying source code, running verification checks, and performing Git operations (commit/push/deploy) when authorized.
- Sanitization: There are no mechanisms described for validating, filtering, or sanitizing the content of the external reports before they are used to generate code changes.
Audit Metadata