tlm provides local, offline command-line assistance powered by open-source LLMs via Ollama. It targets developers and power users seeking an alternative to cloud-based AI assistants, offering features like command suggestion, explanation, and context-aware Q&A without requiring API keys or internet connectivity.

How It Works

tlm integrates with Ollama to leverage various open-source models (e.g., Llama 3, Phi-4, DeepSeek-R1, Qwen) directly on the user's machine. It supports automatic shell detection for seamless integration and offers a Retrieval Augmented Generation (RAG) capability for context-aware queries, allowing users to provide local file paths or patterns for more relevant responses.

Quick Start & Requirements

• Installation: Recommended via script (curl ... | sudo -E bash for Linux/macOS, Invoke-RestMethod ... | Invoke-Expression for Windows PowerShell) or go install github.com/yusufcanb/tlm@1.2 if Go 1.22+ is installed.
• Prerequisites: Ollama must be installed and running locally.
• Resources: Requires sufficient local resources to run LLMs.
• Docs: Usage examples are provided within the README.

Highlighted Details

• No API keys or internet connection required.
• Supports automatic shell detection (Bash, Zsh, PowerShell).
• Offers one-liner command generation and explanation.
• Features "no-brainer RAG" for context-aware Q&A.
• Allows experimentation with various Ollama-compatible models and parameters.

Maintenance & Community

The project is maintained by yusufcanb. No specific community channels or roadmap links are provided in the README.

Licensing & Compatibility

The README does not explicitly state a license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The ask command is marked as beta. The README does not detail specific model performance benchmarks or provide explicit compatibility information for all operating systems and hardware configurations beyond general support for macOS, Linux, and Windows.