Contributing¶

To contribute to this tool, first checkout the code. Then run the tests with uv run:

cd llm
uv run pytest

You can run your development copy of llm using uv run as well:

uv run llm --help

Updating recorded HTTP API interactions and associated snapshots¶

This project uses pytest-recording to record OpenAI API responses for some of the tests, and syrupy to capture snapshots of their results.

If you add a new test that calls the API you can capture the API response and snapshot like this:

PYTEST_OPENAI_API_KEY="$(llm keys get openai)" uv run pytest --record-mode once --snapshot-update

Then review the new snapshots in tests/__snapshots__/ to make sure they look correct.

The default OpenAI plugin has a debugging mechanism for showing the exact requests and responses that were sent to the OpenAI API.

Set the LLM_OPENAI_SHOW_RESPONSES environment variable like this:

LLM_OPENAI_SHOW_RESPONSES=1 uv run llm -m chatgpt 'three word slogan for an an otter-run bakery'

This will output details of the API requests and responses to the console.

Use --no-stream to see a more readable version of the body that avoids streaming the response:

LLM_OPENAI_SHOW_RESPONSES=1 uv run llm -m chatgpt --no-stream \
  'three word slogan for an an otter-run bakery'

Documentation for this project uses MyST - it is written in Markdown and rendered using Sphinx.

To build the documentation locally, run the following:

just docs

This will start a live preview server, using sphinx-autobuild.

The CLI --help examples in the documentation are managed using Cog. Update those files like this:

just cog

You'll need Just installed to run these commands.

To release a new version:

Update docs/changelog.md with the new changes.
Update the version number in pyproject.toml
Run just cog to update docs/fragments.md with the new version number.
Create a GitHub release for the new version.
Wait for the package to push to PyPI and then...
Run the regenerate.yaml workflow to update the Homebrew tap to the latest version.