read

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

• Third-party content exposure detected (low risk: 0.10). The agent skill fetches arbitrary external URLs and PDFs provided by the user, ingesting outsider-authored web content via the read workflow.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill's scripts/fetch.sh script executes runtime HTTP requests to third-party services (https://defuddle.md/ and https://r.jina.ai/) when proxy extraction is enabled, acting as an operational dependency with unknown/third-party provenance.