Ai Bible

What is Ai Bible about?

Ai Bible offers a model‑context‑protocol (MCP) server that delivers precise Bible verse retrieval for LLMs, along with a Dockerized service that mimics the OpenAI completions endpoint. It aims to make biblical data accessible to AI systems in a reliable, repeatable manner for scholarly and teaching purposes.

How to use Ai Bible?

1. Build the MCP server – run the build steps defined in the repository (e.g., npm run build if a script exists).
2. Run the Docker container:

   docker build -f completions/Dockerfile -t mcp-server .
   docker run -p 8002:8000 mcp-server
   

3. Access the Swagger UI at http://localhost:8002/docs to test the get-verse endpoint.
4. Configure your LLM client (e.g., Claude Desktop) to point to the mcp-server.stdio.js file located in the build folder.

Key Features of Ai Bible

• MCP server for deterministic verse lookup across LLMs.
• Docker container exposing an OpenAI‑compatible completions API.
• Swagger documentation for quick testing.
• Web front‑end available at http://ai-bible.com for pocket‑Bible access.
• Open source under GNU GPL v3.

Use Cases of Ai Bible

• Academic research requiring consistent biblical citations from AI models.
• Educational tools that generate explanations or commentary on specific verses.
• Integration with custom LLM pipelines (Claude, Ollama, etc.) for theological chatbots.
• Local development environments that need an OpenAI‑style endpoint without external network calls.

FAQ from the Ai Bible

Q: Do I need an OpenAI API key? A: No, the Docker container provides a self‑hosted completions endpoint that works offline.

Q: Which languages are supported? A: The API currently accepts english as the language parameter; additional languages can be added via data files.

Q: Can I use other LLMs besides Claude? A: Yes, any model that can call an MCP server or the OpenAI‑compatible endpoint can be configured.

Q: How is the data licensed? A: The source code is GPL v3; biblical text files have their own licenses detailed in LICENCE.md.

Q: Where can I see the API documentation? A: Visit http://localhost:8002/docs after starting the Docker container.

ai-Bible

ai-Bible is a project that explores the use of AI within a context of interpreting and understanding biblical text. This repository contains mcp-servers and a container for compatibility with the openai completions API that support an AI or Large Language Model reliably and repeatably lookup data so that it can be represented in different forms for research or educational purposes with some confidence that results will be reproducable and reasonable.

For web accessible front end as a pocket bible see http://ai-bible.com

mcp-server for Claude etc

The mcp-server contains the current implementation of a server for repeatedly and reliably retrieving bible verses when using LLMs. Claude Desktop can be configured to use the mcp-server.stdio.js file built in the build folder of this project as an mcp-server.

See the README.md in that subfolder for detailed information.

docker-container for completions

The docker container wraps the mcp server up using mcpo in order to turn it into server supporting the openai completions api. Run these commands from the project root after building the mcp-server.

docker build -f completions/Dockerfile -t mcp-server .
docker run -p 8002:8000 mcp-server

You can check it is running be checking the swagger api page:

http://localhost:8002/docs

Try the get-verse api with parameters:
{
  "reference": ["Gen.1.1", "Gen.2.1"],
  "language": "english"
}

One way to access the completions api is via Open WebUI and then you can do everything locally with a LLM via Ollama with a model such as llama 3.1 8b, see:

https://docs.openwebui.com/getting-started/quick-start/

Contributing

Contributions are welcome! Please feel free to submit a pull request or open an issue for any enhancements or bug fixes.

License

This project source code is under the GNU GPL v3 Licence. Within the project there are data files that come under different licences. See the file LICENCE.md for details of the GPL licence.