Skip to content

Lazurio documentation

Explore Lazurio. From your first steps and company modules to security and operations.

Lazurio is a working environment for people and AI agents. It keeps company procedures, plans and applications where the team can find them and review changes.

It builds on GitHub, which can hold tasks, documentation and decisions as well as code. Version history shows who changed what. Permissions control who can read, edit and approve the work. You do not need to move accounting, email or every other existing service into GitHub.

Company work is translated by Lazurio into private GitHub processes.
The business remains more than GitHub. Lazurio brings into it the work that benefits from versioning, access control and review.

The agent works with the materials and tools available for the task. It prepares a draft, checks it and presents it for review. You define the goal; an authorized person decides whether to publish or deploy the result.

People direct the work, Task Agents prepare Drafts, people review them, and authorized publication creates shared Organization truth.
GitHub is the context for the whole loop, not a step in it. People direct and authorize; Task Agents prepare the editable work between those decisions.

What Lazurio is today

The version described in the linked sources runs from source with Git and Bun. It includes Launchpad, diagnostics, an experimental CLI v0 and the operating contracts used by connected Organizations and modules. A simple packaged installation is a target, not a released distribution.

You choose the AI model and client separately. Lazurio itself is not a managed SaaS service or a universal security sandbox for AI tools. Review the selected providers’ terms before deployment. The security and control evidence shows where those claims come from and what they do not prove.

Browse the documentation

If you are new to Lazurio, open the Guide. The other sections cover technical requirements, access and operations.

Lazurio is currently source-available under FSL-1.1-Apache-2.0. See the security and control evidence for the license, maturity and assurance details relevant to an approval decision.