Skip to content
UsecaseBoard
← All use cases
githubsports coaching

Turn rec basketball footage into clips, drills, and checkable fixes

At a glance

A recreational player uses a coding agent to review phone footage and return a coaching page that links each correction to a clip and next-practice target. This workflow is listed under sports coaching, video analysis, basketball, practice.

SourceBuilding a tool for these workflows? See weekly rates.

What they did

1) Have a local sports video, Python 3.11 through uv, ffmpeg, and a coding agent that can run shell commands and inspect images. Install the CLI with `uv tool install --python 3.11 git+https://github.com/mariusDing/sports-video-coach`. 2) Ask the agent to read the repository's AGENTS.md and run `sports-video-coach run <video> --sport basketball`. Whenever the CLI exits with code 3, follow the named guide, write the requested review or report file, and run it again until it exits with code 0. 3) Open the finished coaching page and review the clips, a drill for each fix, and a target to check at the next practice; optionally use the rendered explained video.

Worth knowing

The project is alpha. Frame sheets are sampled about every 0.5 to 2 seconds by default, so fast details can be missed; frames read by the agent go to its model provider. The discussion author built it for adult rec basketball footage.

Save ideas in this browser. Upvotes count across the board.

Prompt excerpt

No prompt excerpt was provided by the source.

Sponsored

CodeBuddy

Tencent Cloud AI coding assistant for IDE and CLI.

Try CodeBuddy ↗
Explore all →