find-skills

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The document is benign documentation for finding/installing skills but it instructs bypassing user confirmations (e.g., using -g -y) which weakens a safety boundary and is suspicious without clear malicious intent.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (low risk: 0.10). In SKILL.md, the runtime workflow for “find-skills” instructs the user-facing CLI to accept a free-form query (e.g., npx skills find [query]) and search the open skills ecosystem, which can cause the agent to ingest outsider-authored text from the ecosystem results.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill includes runtime install commands that fetch and execute remote skill code (e.g., "npx skills add vercel-labs/agent-skills@react-best-practices"), so external content would be fetched at runtime and could execute code or control agent behavior.