linkedin-scheduling

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the author's MCP tool package via the npm registry to enable data extraction features.
  • Evidence: The command npx -y @coldiq/mcp@latest is documented in SKILL.md as the method to connect the environment to the author's API.
  • [PROMPT_INJECTION]: The skill utilizes tools that ingest data from external LinkedIn posts, influencer profiles, and live signals, which creates a surface for indirect prompt injection (Category 8).
  • Ingestion points: Tools mcp__coldiq__extract_post_engagement, mcp__coldiq__find_influencers, and mcp__coldiq__find_signals as described in SKILL.md.
  • Boundary markers: Absent. There are no instructions or delimiters provided to the agent to ignore potentially malicious instructions embedded in the external content.
  • Capability inventory: The ingested data is used to inform the agent's recommendations for posting times, frequency, and content velocity strategies described in SKILL.md.
  • Sanitization: Absent. The skill instructions do not mention any filtering, validation, or escaping of the external data ingested by the tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 06:37 PM
Security Audit — agent-trust-hub — linkedin-scheduling