web-search

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external data provided by the user (search results) as described in the 'Evaluate Results' section of SKILL.md.
  • Ingestion points: User-provided search results shared during the evaluation phase.
  • Boundary markers: The instructions lack explicit delimiters or instructions to ignore embedded commands within the analyzed search results.
  • Capability inventory: The skill is configured to use the 'create_docx' tool (as seen in the 'office-mcp' frontmatter), which allows for writing to the file system.
  • Sanitization: There are no instructions for sanitizing or escaping content before analysis or document creation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 07:54 PM
Security Audit — agent-trust-hub — web-search