The Agent Skills Directory

[Unverifiable Dependencies] (HIGH): The script fetch.py executes npx -y @smithery/mcp-fetch, which downloads and runs code from the npm registry at runtime without version pinning or source verification.
[Indirect Prompt Injection] (HIGH): The skill fetches and processes untrusted external content. 1. Ingestion points: The fetch_url method in fetch.py retrieves data from arbitrary URLs. 2. Boundary markers: None; the content is returned directly to the agent's context. 3. Capability inventory: The script uses subprocess.Popen and performs network requests. 4. Sanitization: No sanitization or filtering is performed on the retrieved content.
[Command Execution] (MEDIUM): The script uses subprocess.Popen to execute system commands, which relies on the local npx environment and allows for potential execution of unverified code.

fetch