replica

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

• Third-party content exposure detected (high risk: 0.75). High: the required runtime workflow runs Playwright-based diff probes against a user-supplied live URL (public web content), and those scripts navigate and extract DOM/text for content-diff.mjs/visual-diff.mjs, which can include outsider-authored free text (e.g., page content, cookie/consent/marketing interstitial text) into the LLM context via the agent’s tool outputs/logs.