qingningLime/Cliptolution

Leverages DeepSeek LLM with a modular tool ecosystem (Faster-Whisper, Ollama Qwen vision models, FFmpeg) to analyze video/audio content through multi-step pipelines—extracting speech, analyzing visuals, and generating structured reports—before executing editing actions via integrated tools. Supports interactive creative workflows where the agent autonomously plans and executes complex video tasks (subtitling, creative scripting, music analysis) through natural language prompts, with configurable memory systems and chainable tool execution up to 15 steps.

No commits in the last 6 months.