cli-hub-meta-skill

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches a "Live Catalog" of available tools from a remote DigitalOcean space (reeceyang.sgp1.cdn.digitaloceanspaces.com). DigitalOcean is a well-known service; the download is documented as part of the skill's primary function to provide up-to-date tool listings.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from a remote Markdown file, which creates a surface where malicious instructions could potentially influence agent behavior.
  • Ingestion points: Remote catalog file at https://reeceyang.sgp1.cdn.digitaloceanspaces.com/SKILL.md.
  • Boundary markers: None identified in the skill instructions to delimit external content.
  • Capability inventory: Package installation via pip and execution of installed CLI tools.
  • Sanitization: No evidence of sanitization or validation of the remote catalog content before it is processed by the agent.
  • [DYNAMIC_EXECUTION]: The skill facilitates the dynamic installation and execution of Python packages (cli-anything-*) via the cli-hub install command, which serves as a wrapper for pip operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:04 PM
Security Audit — agent-trust-hub — cli-hub-meta-skill