Context7 - Up to date LLM answers // Amro Elhusseini

Introduction
Understanding Local LLMs
Email RAG Explained
Setting up the Environment
Implementing Local LLM for Email RAG
Challenges and Solutions
Conclusion

Introduction

To understand basic LLM stuff, the models gets trained on a set of data to a particular date, and the data used for training is usally not everything in the whole world,

So for example, in my case, Cisco has a lot of OS models , with many Command versions, usually whenever i ask Google’s gemin, it shares to me a command that actually not suitable for my version of the OS,

In that case, i need some way to fetch the up to date documentation for the LLM to consuklt for my question.

Context7 is a tool (MCP server) service that provides your LLM with the up to date documentation for many libraries in case you are asking fort a command or a class/funstion for your code generation.

Many times, whenever you ask an LLM to generate a command or code, it will generate something random if it does not find the answer in its trained data or using its search tools.

Context7 handles the part where it will get the correct documentation for the AI model to consult.

For us network engineers for example, you can find Context 7 support for our Command lines, and scripting libraries, being kept up to date:

You have to 1st sign up for Context7 API key for free. @ https://context7.com/.

then its easy peasy as follows.

Understanding basic underlying technologies in LLM

Setup

We need to clone the Context7 Repo, to build the docker image : git clone https://github.com/upstash/context7.git
Now that we clone the Dockerfile, run : docker build -t context7-mcp .

edit the Gemini’s Settings.json, or the coresponding configuration file , for Claude desktop or OpenAI’s Codex

$ vim  ~/.gemini/settings.json
{
  "mcpServers": {
    "Сontext7": {
      "autoApprove": [],
      "disabled": false,
      "timeout": 60,
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "CONTEXT7_API_KEY=YOUR-CONTEXT7-API-KEY",
        "context7-mcp"
      ],
      "transportType": "stdio"
    }
  },
  "selectedAuthType": "gemini-api-key",
  "theme": "Default"
}

References

kotaemon , An open-source clean & customizable RAG UI for chatting with your documents. Built with both end users and developers in mind. https://github.com/Cinnamon/kotaemon
Llama.cpp Tutorial: A Complete Guide to Efficient LLM Inference and Implementation https://www.datacamp.com/tutorial/llama-cpp-tutorial
GGUF versus GGML https://www.ibm.com/think/topics/gguf-versus-ggml
Data Is Fueling the AI Revolution. What Happens When It Runs Out? https://alumni.berkeley.edu/california-magazine/online/data-is-fueling-the-ai-revolution-what-happens-when-it-runs-out/

Table of Contents

Introduction

Understanding basic underlying technologies in LLM

Setup

References