Our mission
Help people and AI do useful work together. Give them the knowledge to understand the job, the tools to act, and the infrastructure to put the result to use.
The BitGarten platform
Start with what you know. Give agents the context and tools to help. Keep the useful results, so the next project starts a little further ahead. That’s how your BitGarten grows.
Help people and AI do useful work together. Give them the knowledge to understand the job, the tools to act, and the infrastructure to put the result to use.
We learn through our projects, conversations, decisions, and mistakes. Our vision is for that knowledge to stay useful: available to the right people and agents, wherever the next piece of work happens.
We built BitGarten after seeing the same patterns across companies and startups: people choosing their own AI tools, working locally, and needing to share their work and contribute to a common context. Information kept collecting in isolated tools and conversations.
BitGarten keeps the channels open. People choose their tools; shared knowledge flows between files, agents, and systems. A Markdown and OKF foundation, local sync, shared skills, guidelines, and managed memory let that context keep growing.
Find the brief. Create the document. Build the application. Keep it running. The workspace and its built-in cloud platform carry context through each step.
Markdown, OKF, Drive files, documents, and Omni memory keep your context rooted in sources. Semantic search and a knowledge graph help people and agents find what matters.
Use Codex, Claude Code, your own agents, and familiar tools on local files and cloud environments. Connect systems through MCP, CLI, APIs, connectors, shared skills, and extensions.
The doc-engine, artifacts, assets, collaboration, and comments bring people and generated work together. Keep sources, feedback, and revisions close to the result.
Host apps and managed databases, store assets through S3, and use a built-in IdP for workspace and app access. Mutagen, pipelines, and app monitoring connect local changes to checked releases and ongoing operations.
Built for an open AI ecosystem
Work with BitGarten directly, from the filesystem with Codex and Claude Code, or with ChatGPT Work and Claude Cowork. MCP, CLI, APIs, shared skills, and extensions carry shared context and useful tools into the place you choose to work.
Update a guide. Save the reasoning behind a decision. Keep a useful skill. These small acts nurture the context that people and agents draw on. A proposal, a report, or an app can grow from the same well-tended knowledge, tools, and infrastructure.
Company identity and explicit resource rights give people and agents what they need. Sandboxed agent environments, scoped tools, and approval boundaries keep that access deliberate.
A useful note today. A better starting point tomorrow.
Open workspace