Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
74 lines (50 loc) · 2.08 KB

CONTRIBUTING.md

File metadata and controls

74 lines (50 loc) · 2.08 KB
Raw

Contributing to tuto

Thanks for helping improve tuto.

The most valuable contributions are real tutorial failure cases. If a video contains a setting, button, number, or operation that tuto misses or misreads, please report it with enough evidence to reproduce the problem.

Good bug reports

Please include:

  • a public YouTube URL
  • the approximate timestamp
  • the expected result
  • the actual result
  • your operating system
  • Python, Claude Code, and tuto versions when relevant

Do not include API keys, credentials, private videos, or confidential material.

The tutorial failure template is the best starting point when the problem is tied to a specific video.

Development setup

The Python tests require Python 3.11+ and pytest. Tests that exercise real frame extraction also use ffmpeg/ffprobe; they are skipped when ffmpeg is not available.

Install the runtime tools using the commands appropriate for your system:

pip install -U yt-dlp pytest
# Windows
winget install Gyan.FFmpeg

# macOS
brew install ffmpeg

Run the regression suite from the repository root:

python -m pytest tests -q

If Claude Code is installed, you can also validate the plugin manifest locally:

claude plugin validate .

Pull requests

  1. Keep changes focused.
  2. Add or update tests for behavioral changes.
  3. Run the existing test suite before opening a pull request.
  4. Explain the failure mode your change fixes.
  5. Prefer measurable improvements over prompt-only claims.

Changes that alter extraction, auditing, or orchestration behavior should ideally include a regression case. Documentation-only changes should still verify every referenced command and repository path.

Useful contribution areas

  • macOS and Linux validation
  • caption/screen disagreement cases
  • missed-step detection
  • audit false positives
  • performance regressions
  • token-cost regressions
  • documentation and installation fixes

Design principle

When evidence is unclear, prefer an explicit unresolved result over a confident guess.