byted-bytehouse-cluster-diagnostics

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/diagnostics.py, the required workflow takes outsider-provided free text from sys.argv[1] (question) and sends it verbatim as payload["input"] to the ByteHouse Matrix Agent API, then ingests streamed data: ... message content (event_data.message.content) and prints it.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime script (scripts/diagnostics.py) sends the user's question to a remote Matrix Agent API endpoint constructed as https://{host}/matrix/v1/conversation (using the BYTEHOUSE_HOST environment variable), so the content returned by that external URL directly controls the diagnostics output shown to the user.