url-to-markdown

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

• Third-party content exposure detected (high risk: 0.95). The required workflow calls a hosted pagewatch service to fetch and extract the user-supplied URL’s content (outsider-authored free text) and then returns it as body["markdown"] on stdout for the agent to ingest into LLM context; see the /v1/read path in pagewatch_client.py.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill makes runtime calls to the hosted pagewatch service at https://pagelens.dev (BASE_URL) and prints server-provided fields like "human_message" and verification links verbatim, meaning remote responses can directly control the agent's printed instructions.