gpt-orchestration-auto
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides high-level instructions and configuration for an AI agent to manage long-running tasks. There is no executable code, network usage, or sensitive file access defined in the provided files.
- [INDIRECT_PROMPT_INJECTION]: The skill workflow involves researching external information and integrating returned artifacts, which creates a surface for indirect prompt injection. The skill includes instructions to verify behavior and release evidence to mitigate this risk.
- Ingestion points: External research findings and artifacts integrated into the project (SKILL.md).
- Boundary markers: Instructions specify verification of "accepted behavior" and reconciliation of drift.
- Capability inventory: No subprocess calls, network operations, or file-write commands are implemented in the skill files.
- Sanitization: The process relies on explicit verification stages and reconciled findings before implementation.
Audit Metadata