Obot is an open-source AI governance platform from Obot AI that helps organisations manage, secure and govern their entire AI ecosystem. The GitHub project has about 1,100 stars and is in very active development, with updates landing in October 2026.
As companies adopt AI chat clients, MCP servers and agents, IT teams need one place to control what is connected, who can use it and which credentials are involved. Obot provides that shared infrastructure: it connects AI clients to approved models and tools, distributes vetted MCP servers and skills, manages agent access and credentials, and can run hosted agents and chat experiences. Recent capabilities include running Claude cloud sessions on an organisation’s own Kubernetes infrastructure, so users keep working in Claude while Obot governs the environment underneath. Identity and git integrations (GitHub, GitLab, Bitbucket) tie AI usage into existing enterprise workflows, and the whole platform is a single open-source codebase that can be self-hosted, including in air-gapped setups.
Obot is developed by the team behind several well-known open infrastructure projects, and its development pace is high, with new governance features landing frequently. The documentation covers deployment patterns from a single Docker container to production Kubernetes installations.
Obot is open source under the MIT license: the Community edition is free to self-host, and Obot Enterprise runs the same codebase with commercial packaging and support for organisations that need it. See obot.ai for current enterprise terms.
Platform, IT and security teams at organisations rolling out AI assistants and agents who need central control over models, tools, access and credentials without locking users out of the tools they like.