ip-clause-review

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it is designed to ingest and analyze untrusted external legal documents. Ingestion points: processes content from user-provided file paths, cloud drive links (e.g., Google Docs), and raw pasted text. Boundary markers: absent; the instructions do not define delimiters for processed agreements. Capability inventory: reads and writes to local file system paths within the ~/.claude/plugins/config/ directory and uses MCP tools for legal research. Sanitization: absent; no validation or escaping of document content is performed before analysis.
  • [EXTERNAL_DOWNLOADS]: The skill instructions support fetching content from remote sources like cloud drive links (e.g., Google Docs). As this targets a well-known service, it is documented neutrally, though it serves as the entry point for untrusted data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 01:02 AM
Security Audit — agent-trust-hub — ip-clause-review