find-loops

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime flow involves searching/browsing the public loops directory (e.g., npx agenticloops find / browsing agenticloops.dev), which can fetch outsider-authored free text from public web content that may be ingested into the agent’s LLM context for summarization/recommendation.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs running the CLI via "npx agenticloops", which at runtime fetches and executes remote code from the npm registry (e.g. https://registry.npmjs.org/agenticloops), so an external package controls installation/execution of loops.