Cabinet provides a local, open-source knowledge base for large language models, allowing users to integrate diverse data sources directly on their system. It accepts various file types, including CSVs and PDFs, alongside inline web applications, creating a rich data environment for AI agents. Designed primarily for developers and technical users, Cabinet offers full control over your data and AI interactions without relying on external cloud services. It integrates with Claude Code, enabling the deployment of custom agents complete with features like heartbeats and job management, significantly extending the capabilities of your LLMs. This setup ensures that your AI can access and process proprietary information securely and privately. Ultimately, Cabinet helps users build more intelligent, context-aware AI applications by giving their LLMs a private, extensible data foundation for complex tasks.