website-visitors

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install and connect the vendor's official MCP server using the command npx -y @coldiq/mcp@latest. This is a standard installation procedure for the author's own tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external website visitor signals via tools such as mcp__coldiq__find_signals and mcp__coldiq__find_people. This introduces a potential surface for indirect prompt injection if the ingested visitor data contains malicious instructions.
  • Ingestion points: External visitor signals and person data retrieved from identified tools in SKILL.md.
  • Boundary markers: Absent; there are no specific delimiters or warnings to ignore instructions within the retrieved data.
  • Capability inventory: The skill uses tools to research, enrich, and process contact information based on external signals.
  • Sanitization: No explicit sanitization or filtering of the incoming signal data is described in the provided instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 08:55 AM
Security Audit — agent-trust-hub — website-visitors