First Tool: Context Gatherer for LLM Interactions

[Quasimondo]

Context Gatherer for LLM Interactions

The first development tool for OpenMender has been added: A context gatherer that helps contributors interact with Large Language Models (LLMs) by providing structured repository information.

What It Does

The context gatherer extracts relevant information from the OpenMender repository and formats it for optimal LLM consumption. This includes:

• Repository structure and documentation
• Open issues
• Discussions
• Contributing guidelines
• Component-specific information

How to Use It

1. Clone the repository
2. Install requirements:

pip install -r bootstrap/tools/requirements.txt

3. Run the tool:

python bootstrap/tools/context_gatherer.py --context basic

For better results, use with a GitHub token:

python bootstrap/tools/context_gatherer.py --token YOUR_GITHUB_TOKEN --context all

Use Cases

1. Getting Started
   Feed the context to an LLM to understand the project structure and find good entry points for contribution.

2. Issue Analysis
   Gather current issues and their context to help identify patterns or potential solutions.

3. Documentation Work
   Get a comprehensive view of existing documentation to suggest improvements or additions.

Next Steps

The tool is open for improvements. Some ideas:

• Add more context types
• Improve formatting for specific LLMs
• Add filtering options
• Create integration examples for popular LLM APIs

Check out the tool documentation for more details.

---

Share your experiences or suggestions for the tool in the comments below.

[Quasimondo]

Added retrieval of the repository's file tree and the option to get the source code of particular files. Updated docs here: tool documentation