competitor-signals

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and encourages the use of an external package @coldiq/mcp which is downloaded from the npm registry.
  • [REMOTE_CODE_EXECUTION]: The instructions include a command to execute remote code: npx -y @coldiq/mcp@latest. This pattern downloads and runs the latest version of the vendor's MCP server directly in the shell environment.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it is designed to ingest and process untrusted data from external platforms.
  • Ingestion points: The skill scrapes content from G2, Capterra, and LinkedIn company pages, including reviewer comments and profile descriptions.
  • Boundary markers: There are no explicit instructions or delimiters provided to ensure the agent treats the scraped content strictly as data rather than instructions.
  • Capability inventory: The skill uses tools to enrich contacts and generate outreach based on the scraped content (e.g., mcp__coldiq__enrich_person).
  • Sanitization: No sanitization, filtering, or validation steps are described for the external data before it is incorporated into the agent's workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 06:37 PM
Security Audit — agent-trust-hub — competitor-signals