commercial-skills

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external documents from the workspace, which could contain adversarial instructions aimed at biasing the agent's commercial recommendations.
  • Ingestion points: The skill explicitly searches for and ingests RFP PDFs, pipeline exports (CSV), and MSA redlines within the user workspace.
  • Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands provided in the skill text to protect the processing of these documents.
  • Capability inventory: The orchestrator performs file discovery, routes data to sub-skills, and generates summarized commercial digests; it does not utilize network or shell tools in the analyzed code.
  • Sanitization: The instructions lack defined mechanisms for validating or sanitizing the content extracted from external business documents before it is processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 05:14 AM
Security Audit — agent-trust-hub — commercial-skills