explore-category

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs standard network requests to the vendor's own domain (agenticstack.sh) to fetch JSON metadata about tool categories. This behavior is consistent with its stated purpose and does not involve sensitive user data or privileged system access.
  • [PROMPT_INJECTION]: The skill ingests editorial markdown and metadata from an external API response, creating a surface for indirect prompt injection.
  • Ingestion points: JSON response data from https://agenticstack.sh/api/json/categories/{slug} (specifically the body and feature_definitions fields).
  • Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when presenting the external content to the agent.
  • Capability inventory: The skill does not request or use high-risk tools (e.g., shell access, file writes, or credential access).
  • Sanitization: No sanitization or validation of the retrieved editorial content is specified in the instruction file.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 03:27 PM
Security Audit — agent-trust-hub — explore-category