mcp-ffmpeg

# MCP-FFMPEG An MCP (Model Context Protocol) server and CLI for running FFmpeg jobs via a job queue with configurable parallel workers. ## Features - **Job queue** — Enqueue video jobs; workers process them in the background. - **Parallel workers** — Run multiple jobs at once (number set in config). - **Two interfaces** - **CLI** — Interactive menu to pick an action and enter parameters. - **MCP server** — Tools for AI assistants (e.g. Claude Desktop) to enqueue and check jobs. - **Actions** - **Trim** — Cut a segment from a video (start time + duration). - **Change video format** — Convert to another container (e.g. mp4 → mkv) without re-encoding. - **Change resolution** — Convert to another resolution, height and width provided by user. - **Change Subtitle format** — Convert to another subtitle format (e.g. srt -> vtt). - **Extract Audio** - Extract audio from an input video file - **Caching** — Same inputs produce the same job ID; completed jobs are reused unless `force_run` is used. ## Requirements - **Python** 3.13+ - **FFmpeg** — Must be on your system PATH or set via `FFMPEG_PATH` (see [Configuration](#configuration)). ## Installation Choose one of the following: ### Option 1: Install from PyPI ```bash pip install mcp-ffmpeg ``` Or with [uv](https://docs.astral.sh/uv/): ```bash uv add mcp-ffmpeg ``` ### Option 2: Clone the repository ```bash git clone https://github.com/priyanshum143/MCP-FFMPEG.git cd MCP-FFMPEG ``` Then install from the project root: ```bash # With uv uv sync # Or with pip (editable install) pip install -e . ``` ## Configuration - **FFmpeg path** - Default: use `ffmpeg` from system PATH. - Optional: set env var `FFMPEG_PATH` to the full path of the FFmpeg executable (e.g. for Claude Desktop). - **Worker and paths** Edit `src/MCP_ffmpeg/utils/variables.py` (class `CommonVariables`): - `PARALLEL_EXECUTIONS_ALLOWED` — Number of jobs that can run at once (default: 3). - `WORKER_RE_RUN_TIME` — Sec…