nextjs-modal-integration

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow for streaming LLM responses ingests OUTSIDER-authored free text: the user’s prompt comes from the Next.js client request body (app/api/chat/route.ts → fetch to Modal /generate-stream), and that prompt is then passed into the Modal FastAPI endpoint (modal_backend/llm.py → generate_stream_endpoint(req.prompt)), where it becomes LLM input text.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill makes runtime calls to external Modal endpoints (e.g., https://your-workspace--nextjs-backend-api.modal.run and other MODAL_* URLs such as MODAL_IMAGE_GEN_URL, MODAL_JOBS_URL, MODAL_FILES_URL, MODAL_LLM_URL) which execute remote code/services (image generation, LLM streaming, job processing) and are required for the skill to function.