motion-graphics

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). High-risk: the code includes a deliberate "auto" path that base64-encodes images and POSTs them (with an API key) to an external LLM endpoint (data/credential exposure), plus headless browser code that loads remote JS/tile endpoints and runs with --no-sandbox, all of which enable intentional external data transfer and remote dependencies.

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

• Third-party content exposure detected (medium risk: 0.30). At runtime, the workflow may read outsider-authored free text from user-supplied webpages/news/tweets via the search-driven Step 2 — Source path (shot-plan.json.asset_needs → phases/source/resolve.mjs → media-use capture/resolve), then uses the extracted text fields (e.g., tweet/news headline/keyword) in later build steps.

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

• Potentially malicious external URL detected (high risk: 0.90). The basemap bake helper injects and executes third‑party scripts and tiles at runtime (e.g. https://cdn.jsdelivr.net/npm/maplibre-gl@5.24.0/dist/maplibre-gl.js and the Esri/CARTO tile URLs such as https://server.arcgisonline.com/ArcGIS/rest/services/World_Imagery/MapServer/tile/{z}/{y}/{x}), which are fetched and executed during bake-basemap.mjs runtime to control rendering and geometry export.