The Agent Skills Directory

[SAFE]: The skill operates as a high-level orchestrator for a video creation pipeline, managing workflow state through local file system organization and delegating technical tasks to modular sub-skills.
[EXTERNAL_DOWNLOADS]: The skill references and utilizes content from established media platforms including YouTube, Bilibili, TikTok, and Xiaohongshu. It also integrates with official generative AI APIs from ByteDance (Seedance 2.0). These integrations are legitimate and align with the skill's primary function of media production.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it ingests and processes untrusted data from external video sources to generate scripts and summaries. This risk is inherent to its operational purpose and is managed by the agent's inherent safety guardrails.
Ingestion points: Video transcripts and reference metadata retrieved from external social media URLs.
Boundary markers: Absent. The instructions do not specify explicit delimiters or warnings to ignore instructions within the retrieved content.
Capability inventory: The workflow involves file system writes for project management and the invocation of external rendering tools (e.g., Remotion, Hyperframes) and APIs.
Sanitization: No specific content sanitization or filtering logic is defined within the orchestration instructions.

space-video