Releases: varadanvk/llm-cli
Releases · varadanvk/llm-cli
v0.0.7
v0.0.7 Update
New: OpenRouter Support
- Added OpenRouter API with access to Gemini 2.5 Pro, Grok-3, Grok-4
- Select "custom" to use any OpenRouter model
Easier Setup
- Load API keys from
.envfile automatically - Better error messages and setup feedback
Install/Update
pip install --upgrade llm-chat-cli
lmci setupAdd these to your .env file:
OPENROUTER_API_KEY=your_key
GROQ_API_KEY=your_key
OPENAI_API_KEY=your_key
ANTHROPIC_API_KEY=your_key
CEREBRAS_API_KEY=your_key
v0.0.6
What's Changed
🐛 Bug Fixes
- Fixed markdown rendering for code blocks
- Code blocks now properly accumulate content instead of rendering empty blocks
- Added support for inline code highlighting with backticks
- Improved buffering logic for multi-line markdown content (tables, lists, etc.)
🔧 Technical Details
- Simplified the function
- Fixed state machine logic that was causing code to appear outside markdown blocks
- Better handling of streaming chunks across code block boundaries
This release fixes the issue where code blocks would render incorrectly during streaming, especially when chunks were split across code block markers.