clause-analyze

Pass

Audited by Gen Agent Trust Hub on Jul 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing external data without explicit boundary markers.
  • Ingestion points: The skill reads local project documents and fetches content from external websites via WebSearch and WebFetch in Step 1.
  • Boundary markers: There are no instructions provided to treat the content of contracts or web pages as data only, nor are there delimiters used to isolate this content from the agent's instructions.
  • Capability inventory: The agent has access to sensitive tools including Bash, Write, and Read, which could be misused if instructions are successfully injected into the processed data.
  • Sanitization: The skill does not implement validation or sanitization for the data retrieved from external sources before it is analyzed.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 1, 2026, 02:48 PM
Security Audit — agent-trust-hub — clause-analyze