observing-atproto

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

• Third-party content exposure detected (high risk: 0.90). The skill ingests and analyzes public, user-generated ATProtocol/Bluesky content by connecting to the Jetstream firehose (wss://jetstream2.us-east.bsky.network/subscribe) and calling the public API (https://public.api.bsky.app/...), so the agent will read untrusted third-party posts that could carry indirect prompt injections.