account-selection
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides instructions for the user to set up a Model Context Protocol (MCP) server using npx -y @coldiq/mcp@latest. This command downloads and executes code from the npm registry. The package belongs to the skill author ('cold-iq').
- [EXTERNAL_DOWNLOADS]: The skill references and fetches data from the vendor's API at api.coldiq.com and provides links to coldiq.com for service acquisition.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external company, technographic, and CRM data from various sources to perform account scoring, which presents an attack surface for indirect instructions. 1. Ingestion points: Data entering via ColdIQ MCP tools and fallback tools like Apollo or HubSpot mentioned in SKILL.md. 2. Boundary markers: None explicitly defined in the provided instructions to delimit external content. 3. Capability inventory: Uses network operations for API calls and potential file access via skill resources. 4. Sanitization: No explicit sanitization or validation steps are defined within the skill instructions.
Audit Metadata