v1.8.0

What's new

1. lumen explain now also outputs the AI Provider it used to answer the request based on your configuration
2. New provider: DeepSeek 🐋
3. Lumen now ignores certain common files like (package-lock, Cargo.lock) to reduce token usage

1.7.1

FIxes

1. Claude API 400 Request

1.7.0

What's new

1. explain command now supports input from stdin - echo "HEAD" | lumen explain -
2. Packaged for nix
3. Triple dot (merge-base) support for ranged comparisons lumen explain main...featureA
4. Fixes

1.6.0

What's new

Introducing support for ranges 🔅
explain command now supports commits range. Branches can also be compared since branch names are just aliases for their respective HEADs

# lumen explain <from>..<to>
lumen explain HEAD~3..HEAD            # Last 3 commits
lumen explain main..feature/A         # Branch comparison

1.5.0

What's new

1. Added support for OpenRouter
2. Using indoc crate for prompt strings - leads to slightly reduced token usage

1.4.0

What's new

Introducing Advanced Configuration

You can now use a configuration file to configure lumen on a per-project basis.
For now, the only extra option (other than basic ones like provider, modal, api_key) supported is draft.commit_types. This allows for customising your commit types based on your project's need.

For more details, check the "Advanced Configuration" section in the README.md

1.3.0

What's new

You can now ask questions about the entity when using the explain command.

Try it:
lumen explain --diff --query "how will this change affect performance?"
lumen explain HEAD --query "how will you improve this?"

1.2.0

What's new

Added support for Ollama
Reference: https://github.com/ollama/ollama

Try it using:
lumen -p "ollama" -m "your-model" list

1.1.0

What's new

Added --context or -c flag for lumen draft to communicate intent of the change for commit generation.
This effectively allows for more accurate commits at the cost of some (but still) minimal effort.

# Without context:
lumen draft
# Output: "feat(button.tsx): Change button color to blue"

# With context - just add a simple why:
lumen draft --context "match brand guidelines"
# Output: "style(button.tsx): Update button color to align with brand identity"

1.0.0

Lumen is out in 1.0.0!

This doesn't mean that further improvements are not coming, but the initial vision has been accomplished 🎉

What's new

1. Generate git commit message for staged changes using lumen draft - Check README for usage
2. Improved error handling and error reporting.
3. API errors for all providers will now be displayed

There have also been significant improvements to the code for maintainability and future additions