git-courer: git-courer intercepts git operations performed by AI assistants and runs them locally using Ollama, saving thousands of tokens per operation. It adds security checks for secrets and supports auto-configuration of popular AI tools. Releases use Go for version calculation and Ollama for changelog generation.; xLAM: xLAM is a research repository for Large Action Models (LAMs), which aggregates and unifies agent trajectories from diverse environments into a consistent format. It streamlines the creation of a generic data loader optimized for agent training, enabling robust model development across various scenarios.
Automating git commits with AI without token cost
Function calling in LLMs