What problem does it solve? Once OpenStoryline is installed, users still face a multi-step operational burden: configuring model credentials, starting the MCP and Web services, creating sessions, uploading media, and driving multi-turn editing conversations. This Skill automates that entire loop so users can go from raw footage to a rendered video using plain-language instructions. ## Core Features & Use Cases - Service Orchestration: Configures config.toml fields (LLM/VLM credentials, TTS providers, Pexels search) and launches the MCP server and uvicorn Web service as long-running processes with log monitoring. - Session-Based Editing: Creates a session, uploads media, sends editing prompts via a WebSocket bridge script, and preserves the session_id for multi-turn re-editing that produces new output_*.mp4 files. - Output Verification & Delivery: Locates rendered videos in the server cache and can send them to users through Feishu using the bundled sender script. - Use Case: A user asks to "cut a Xiaohongshu-style video" from local clips; the Skill starts the services, creates a session, runs the edit, verifies the output MP4, then applies follow-up requests like "use a more upbeat BGM" on the same session. ## Quick Start Start OpenStoryline and use it to edit my video clips in /path/to/media into a short social-media style video.
