content-engagement

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and instructs the use of the @coldiq/mcp package from the NPM registry to facilitate integration with the Model Context Protocol.
  • [COMMAND_EXECUTION]: Provides a shell command instruction (npx) for the user to initialize the Model Context Protocol server locally using their own API key.
  • [PROMPT_INJECTION]: The skill is designed to process untrusted data from social media platforms, posing a potential surface for indirect prompt injection.
  • Ingestion points: External data enters through LinkedIn post engagement scraping tools (mcp__coldiq__extract_post_engagement).
  • Boundary markers: None provided in the instruction body.
  • Capability inventory: The skill includes external tool invocation via MCP and instructions for local shell execution.
  • Sanitization: No specific content filtering or sanitization measures are described for the social media data ingested.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 06:37 PM
Security Audit — agent-trust-hub — content-engagement