skills/ndpvt-web/arxiv-claude-skills/supchain-bench-benchmarking-real-world-supply/Gen Agent Trust Hub
supchain-bench-benchmarking-real-world-supply
Pass
Audited by Gen Agent Trust Hub on May 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides technical documentation and architectural patterns for building diagnostic supply chain agents. The provided pseudo-code and workflow descriptions are educational and consistent with the stated purpose.\n- [PROMPT_INJECTION]: The skill includes directive language (e.g., 'CRITICAL', 'MUST', 'You MUST call') within suggested system prompt templates. These are instructional for the agent being developed and do not attempt to bypass the host AI's safety guidelines or core instructions.\n- [EXTERNAL_DOWNLOADS]: The content references an academic paper on Arxiv (arxiv.org), a well-known academic repository. This reference is for informational purposes and does not trigger the execution of remote scripts or binaries.
Audit Metadata