wjs-converting-text-to-video

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

• Third-party content exposure detected (high risk: 0.85). Step 1/2 reads <article-folder>/article.md (outsider-authored article text) and Step 2 writes narration_chunks.json, which Step 3 then feeds into Volcano TTS and Step 5/6 uses the resulting narration/timing to drive the HyperFrames composition—so the outsider free text becomes LLM-readable narration content at runtime.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill includes runtime-loaded remote code: the HyperFrames registry ("https://raw.githubusercontent.com/heygen-com/hyperframes/main/registry") is referenced in hyperframes.json and will be fetched by npx hyperframes during lint/inspect/render (pulling component/code at runtime), and the composition index.html loads GSAP from the CDN ("https://cdn.jsdelivr.net/npm/gsap@3.14.2/dist/gsap.min.js"), which is fetched and executed during rendering—both are required runtime dependencies that execute remote code.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)