Engineer with AI agents¶

CLI VS Code Experimental

For teams who want AI assistants to write requirements, architecture and code, while the engine checks every step against the traceability they declared.

Pharaoh, ubCode’s agentic engineer, is Experimental. Try it on a project where you review each change it proposes, and pin your ubc version.

AI agents, checked by the engine says what the engine checks and when, and helps you pick a setup.
Pharaoh Agentic Engineer explains the mental model: AI does the writing, the engine computes what is done and what comes next.
Quick start installs a profile and drives the loop.
Configuring the workflow and Workflow describe your process once.
Reviews and quality analysis and Quality Analysis add reviews that hold a stage until its needs pass.
In VS Code and Chat participant cover Pharaoh’s panels and its chat entry point in VS Code.
Model Context Protocol (MCP) server gives any AI assistant your project’s graph and the product documentation.
Recovering intent from code recovers intent from an existing codebase.
The ubc agent command lists every ubc agent command, the engine behind Pharaoh.
Gate AI-written work gates AI-written work in CI.

Next steps¶

Upgrade path: Pharaoh works on a typed requirements graph; if you do not have one yet, start with Manage requirements and traceability.
For your team: ubCode for teams covers licensing, offline use and support.

Use it at work. ubCode is free for open-source projects, and a private project needs a commercial license. Request a free enterprise trial to evaluate it on your own private project.