GitHub radar
OpenGrok: One command to deploy all AI models on Grok Bot!
An evidence-based wire capture powered model selector for Grok Bot. Select between xAI Grok, Zhipu GLM, Anthropic Claude, Google Gemini, DeepSeek or local llama.cpp models for each agent, ensuring your chosen model works as expected post app update.
OpenGrok is a Python package which provides a model selector for Grok Bot. Each provider mapping has a corresponding wire capture stored in this repo. We follow an ‘evidence or it doesn’t ship’ philosophy. The install command will scan for an existing Grok Bot install and either reuse the bindings or prompt you with 3 questions. It then generates a config file. You can easily try out a binding using the model selector and clicking ‘save’. There is also a doctor utility which takes a baseline of your system during install and keeps track of files and services. Post Grok Bot update it will tell you exactly what changed and what still needs applying. Hot reload provider mappings. Cross platform (Windows/MacOS/Linux). Zero dependencies other than python.
Why a vibe-coder should care
If you are using Grok Bot as your AI code client and you need to change the model it uses (e.g. to Claude, DeepSeek or local) this package will correctly set up the configs and stay working through app updates. No reverse engineering, no manual fiddling, no worrying about wire protocols.
How to install
Copy this and send it to your agent — Claude Code, Codex, any of them:
Install opengrok from https://github.com/OnlyTerp/opengrok following the README — run python setup.py, open the model picker, choose a provider and verify everything is healthy with python tools/doctor.py. Ask me for OpenRouter API keys if needed.
Runs on a regular laptop — only Python needed.
Open on GitHub▌ More finds