pencil-mcp-get-style-guide-tags

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill utilizes 'id' and 'tags' parameters that act as ingestion points for external data into the agent's execution context.
  • Ingestion points: Defined input parameters in SKILL.md (id, tags).
  • Boundary markers: The skill lacks explicit data delimiters but provides a specific logic block for intent verification to ensure the user explicitly requested the Pencil tool.
  • Capability inventory: The skill is scoped to calling the get_style_guide_tags design tool.
  • Sanitization: No specific validation or escaping routines are described for the input parameters.
  • [SAFE]: The skill follows security best practices by including an 'Intent Recognition' section that instructs the agent to verify the user's specific goal before invoking the tool, which serves as a mitigation against unintended triggers or indirect prompt injection attempts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 03:04 PM
Security Audit — agent-trust-hub — pencil-mcp-get-style-guide-tags