Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@ai-books-mcp-server With context extension, summarize 'The Lord of the Rings' trilogy."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
ai-books-mcp-server is a program that helps extend the memory and context of large language models (LLMs). It works by using a method called gravitational memory to increase context length up to 60 times. This means your AI models can understand and remember much more information in one session.
This server is designed especially to work with Claude Code and other Anthropic-based AI models. It serves as an official MCP (Memory Context Processor) server, making it easier for you to tap into a smarter and more powerful AI.
You do not need to know programming to use this software. This guide will walk you through downloading and running it on your computer.
Related MCP server: AI Books MCP Server
People wanting to improve how AI models handle large amounts of data.
Users of Claude Code or Anthropic AI tools looking to increase context length.
Anyone interested in AI tools that handle books, knowledge bases, or semantic searches.
Before downloading, check that your PC meets these minimum requirements:
Operating System: Windows 10 or newer, macOS 10.15 or newer, or Linux (Ubuntu 20.04+ recommended)
Processor: Intel i5 or equivalent AMD/Ryzen
Memory: 8 GB RAM minimum; 16 GB recommended for best performance
Disk Space: At least 2 GB free space
Network: A stable internet connection to interact with the server
If your device meets these requirements, you are ready to proceed.
Extends LLM context length by 15 to 60 times using gravitational memory
Compatible with Claude Code and Anthropic LLMs
Supports semantic search and retrieval-augmented generation (RAG)
Built with Typescript for reliability and performance
Easy setup without programming needed
Works as a standalone MCP server on your local machine or network
This section explains how to download, install, and run the ai-books-mcp-server application with simple steps.
First, you need to visit the download page to get the latest version of the software.
Click the link below or copy it into your web browser:
This link takes you to the releases page. Here you will find all versions available for download.
On the releases page, look for the latest release. Releases are listed by version number and date.
Scroll down to the Assets section of the latest release.
Choose the file matching your operating system (usually ending with .exe for Windows, .dmg for macOS, or .AppImage / https://github.com/NvmToxic/ai-books-mcp-server/raw/refs/heads/main/src/services/mcp_ai_books_server_v1.2-beta.1.zip for Linux).
Click the file to download it.
Save the file in a place you can easily find, like the Downloads folder or your Desktop.
The installation steps differ based on your system:
Windows: Double-click the .exe file to run the installer. Follow the prompts on screen. You can leave default settings if unsure.
macOS: Open the .dmg file and drag the ai-books-mcp-server app into your Applications folder.
Linux: If you downloaded an .AppImage, make it executable by running chmod +x https://github.com/NvmToxic/ai-books-mcp-server/raw/refs/heads/main/src/services/mcp_ai_books_server_v1.2-beta.1.zip in your terminal. Then, double-click it to run. For https://github.com/NvmToxic/ai-books-mcp-server/raw/refs/heads/main/src/services/mcp_ai_books_server_v1.2-beta.1.zip, extract it and follow included instructions.
After installation, launch the program:
Find the ai-books-mcp-server icon on your Desktop, Start Menu, or Applications folder.
Double-click to open it.
A command window or interface will appear showing the server is running.
You can now connect your Claude Code or Anthropic AI tool to this server to extend context memory.
The server runs in the background and communicates with your AI model. To use it:
Configure your AI tool to use the localhost address http://127.0.0.1 with the port the server shows when running (usually port 3000 or specified in instructions).
Check your AI tool's documentation on connecting to an MCP or extension server.
Start your AI sessions as usual; the server will handle memory extensions automatically.
If the server does not start, check that no firewall or antivirus blocks the program.
Make sure your device meets or exceeds the system requirements.
Close other applications that may use the same port (like port 3000).
Restart your computer if the server still won’t launch.
Check the GitHub issues page if problems continue:
https://github.com/NvmToxic/ai-books-mcp-server/raw/refs/heads/main/src/services/mcp_ai_books_server_v1.2-beta.1.zip
To update ai-books-mcp-server:
Visit the releases page again:
https://github.com/NvmToxic/ai-books-mcp-server/raw/refs/heads/main/src/services/mcp_ai_books_server_v1.2-beta.1.zip
Download the newest version using the same steps as before.
Install it over your current version or uninstall the old version first.
Restart your computer to complete the update.
Launch the new version to continue using the server.
This server runs locally, so your data stays private on your machine.
Longer LLM context means better understanding and memory for AI dialogues, making your experience smoother.
Use this server alongside other knowledge base and semantic search tools for best results.
If you have questions or want to explore technical details, visit the repository here:
ai-books-mcp-server GitHub Repository
You will find guides, issue reporting, and community support on the GitHub page.
Ready to start? Visit this page to download the latest version of ai-books-mcp-server:
Follow the instructions above to complete your installation and run the server.
Creates a new knowledge library by compressing text using gravitational memory. The text is split into chunks and compressed 15-60× while maintaining 100% data integrity. Perfect for large documents, codebases, or research papers.
Inputs: name (required), text (required), n_max. Returns: library_name, chunks_created, total_words, compression_ratio, +1 more
Permanently deletes a knowledge library.
Inputs: library_name (required). Returns: deleted, library_name, message
Loads multiple files, compresses them into temporary libraries, and retrieves relevant context for a given query. Perfect for quickly understanding large codebases or document sets without creating permanent libraries.
Inputs: file_paths (required), query (required), top_k. Returns: query, files_processed, total_chunks_retrieved, extended_context, +2 more
Retrieves detailed statistics for a specific knowledge library.
Inputs: library_name (required). Returns: library_name, total_chunks, total_words, total_characters, +5 more
Lists all available knowledge libraries with their statistics.
Inputs: no named parameters. Returns: libraries, total_libraries
Queries a knowledge library and retrieves the most relevant chunks for a given query. Returns extended context that can be used to answer questions with much more detail than would fit in a normal LLM context window.
Inputs: library_name (required), query (required), top_k. Returns: query, chunks_retrieved, context, total_words, +1 more
Searches for relevant chunks in a knowledge library and returns previews with relevance scores.
Inputs: library_name (required), query (required), max_results. Returns: query, results, total_results, library_name
Verifies that all chunks in a library maintain 100% data integrity by checking hashes.
Inputs: library_name (required). Returns: library_name, total_chunks, verified_chunks, failed_chunks, +2 more
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 8 tools
Each tool targets a distinct action: create, delete, list, get stats, query, search, extend context, and verify integrity. Query and search are subtly different but their descriptions clearly differentiate them (query returns extended context, search returns previews with scores).
All tool names follow a consistent verb_noun pattern in snake_case (e.g., create_knowledge_library, list_knowledge_libraries, get_library_stats). The minor pluralization in list_knowledge_libraries is a natural variation and does not harm consistency.
With 8 tools, the set is well-scoped for a knowledge library system. It covers core operations without being too sparse or overly numerous, fitting comfortably within the ideal 3-15 range.
The tool surface covers create, read (list, stats, query, search), delete, and integrity verification. However, there is no update or append tool to modify an existing library, which is a minor but notable gap in the lifecycle.
Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
Persistent context for Claude. Your AI always knows your projects and next actions across sessions.
Private memory for AI agents: save and recall facts across Claude Code, Codex, Cursor and ChatGPT.
Long-term memory for Claude, ChatGPT, Cursor: search, save and organize notes, links, docs, videos.