designmdme-cli

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @crowdlinker/designmdme package from the NPM registry. This is a vendor-provided package used to facilitate the design extraction functionality.
  • [COMMAND_EXECUTION]: The skill executes shell commands including npm, npx, and the designmdme binary to perform authentication, generation, and history inspection.
  • [DATA_EXFILTRATION]: The CLI tool manages account authentication by reading and writing tokens to ~/.designmdme/auth.json and communicates with the vendor domain designmd.me.
  • [PROMPT_INJECTION]: The skill processes untrusted content from external websites to generate design specifications that the agent is then instructed to follow. Ingestion points: External website URLs provided to the CLI tool (SKILL.md). Boundary markers: Absent; there are no instructions to the agent to disregard malicious content or use delimiters for the generated Markdown. Capability inventory: The agent is directed to read the generated file and use its values verbatim when writing UI code. Sanitization: The skill relies on the CLI tool's internal scraping and processing logic to handle external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 01:02 PM
Security Audit — agent-trust-hub — designmdme-cli