define-icp
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs users to install a Model Context Protocol (MCP) server using
npx -y @coldiq/mcp@latest. This is an official tool provided by the skill's author, Cold-IQ, and is intended for legitimate data enrichment functionality. - [DATA_EXFILTRATION]: The skill references the author's API at
api.coldiq.comand provides links to their commercial marketplace. These network interactions are directed to the vendor's own infrastructure to support the skill's core purpose of fetching company intelligence. - [PROMPT_INJECTION]: The skill processes external company and technographic data retrieved through MCP tools, which creates a surface for indirect prompt injection.
- Ingestion points: Data enters the context via
mcp__coldiq__search_companiesandmcp__coldiq__enrich_companytools. - Boundary markers: No explicit delimiters or instructions are used to distinguish external data from internal logic.
- Capability inventory: The skill uses MCP tools for data retrieval and provides scoring based on the output.
- Sanitization: There are no documented steps for validating or sanitizing the external data fetched from the vendor's API.
Audit Metadata