0xsline

@ItsmeAjayKV Kind reminder that community has curated resources for deepseek harness : 0xsline/awesome-deepseek-harness

2/ OpenChatCut — você edita conversando: "corta as pausas" e ele monta a timeline. 0xsline/OpenChatCut

1. Run it (needs Node.js 24): git clone 0xsline/OpenChatCut cd OpenChatCut && npm install cp .env.example .env.local && npm run dev Prefer an installer? macOS, Windows and Linux builds are on GitHub Releases. 2. Open localhost and open your project 3. Add the skill to your agent: npx skills add 0xsline/OpenChatCut

OpenChatCut GitHub Explained: Agent-Native Timelines for Local AI Video Editing openchatcut by 0xsline: 0xsline/openchatcut OpenChatCut is presented as a local-first, agent-native video editor built around a structured multitrack timeline rather than a flattened prompt export. This architecture overview follows the local runtime and browser context, the Model Context Protocol bridge on port 5199, on-demand agent skills, draft overlays with human review, and deterministic rendering through Remotion, WebGL2 GLSL shaders, and FFmpeg. It also covers a cited Apple Silicon M4 Max render comparison: 13 workers took 14.03 seconds for a 4K HEVC test segment, while 4 workers took 8.39 seconds. The closing sections trace CapCut draft handoff, the AGPL-3.0 licensing boundary, and the trade-off between opaque generation and frame-level timeline revision. TimeStamps: 0:00 Black-box outputs and editable timelines 1:07 Local runtime and browser context 2:23 MCP bridge at local port 5199 3:36 Draft overlays and human review 4:41 Deterministic rendering with Remotion and GLSL 5:39 Render worker counts and export timing 6:46 CapCut draft handoff 7:19 AGPL-3.0 licensing boundary 7:43 Why revision needs transparent timelines 🧩 Agent-native multitrack timeline 🔌 MCP bridge and on-demand skills 🎞️ Remotion, GLSL, and FFmpeg rendering ⚙️ Render tuning, CapCut handoff, and AGPL-3.0 scope For developers evaluating local AI video editing, the key distinction is editable structure: timeline layers, reviewable drafts, and deterministic frame placement remain available for later revision. OpenChatCut’s trade-offs center on runtime requirements, render tuning, external finishing, and AGPL-3.0 obligations. #OpenChatCut #AIVideoEditing #GitHub [de-DE] YNAS ynp_7f16942d68110b92 filmisch Privater automatisierter Staging-Upload. Paketidentität: ynp_7f16942d68110b92; Ausgabe: cinematic. [es-US] YNAS ynp_7f16942d68110b92 cinematográfico Carga de entorno de prueba de automatización privada. Identidad del paquete: ynp_7f16942d68110b92; salida: cinematográfica. [fr-FR] YNAS ynp_7f16942d68110b92 cinématique Téléchargement de l'environnement de test d'automatisation privé. Identifiant du package : ynp_7f16942d68110b92 ; sortie : cinématique. [hi] YNAS ynp_7f16942d68110b92 सिनेमैटिक निजी स्वचालन स्टेजिंग अपलोड। पैकेज पहचान: ynp_7f16942d68110b92; आउटपुट: सिनेमाई। [id] YNAS ynp_7f16942d68110b92 sinematik Unggahan staging otomatisasi pribadi. Identitas paket: ynp_7f16942d68110b92; output: sinematik. [it] YNAS ynp_7f16942d68110b92 cinematic Caricamento di automazione privata. Identità del pacchetto: ynp_7f16942d68110b92; output: cinematografico. [pt-BR] YNAS ynp_7f16942d68110b92 cinemático Upload de teste de automação privada. Identificação do pacote: ynp_7f16942d68110b92; saída: cinemática.

OpenChatCut替代ChatCut：AI本地剪片 无需订阅，本地操作。 真实多轨时间轴，AI剪完还能接着调 Agent原生，26个专业技能装进Codex和Claude Code 本地优先，素材和工程都留在自己机器 0xsline/OpenChatCut

@benwvlshvm @ItsmeAjayKV @DeepSeekHarness In case you need some curated plugins :) 0xsline/awesome-deepseek-harness