fw-ai-actions-app

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly requires the agent to perform web_search and mcp_web_fetch to load official third‑party API documentation and "extract" endpoints/auth/body (see rules/ai-actions-api-docs.mdc and SKILL.md "Docs first"), and those fetched public web pages are then used to determine request templates, auth, and implementation—so untrusted external content can directly influence the agent's actions.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill explicitly requires runtime web_search / mcp_web_fetch of third‑party API documentation (blocking implementation) — for example it mandates fetching and using https://developers.freshworks.com/docs/app-sdk/v3.0/common/actions/ to drive action implementation, so external content fetched at runtime directly controls the agent's instructions.