Start building
Faheem Code (فهيم كود) is an AI software engineer for your team. It runs coding agents in sandboxes, connects them to your repositories and issue trackers, and gives you a browser client, a CLI, a Python SDK and a self-hosted deployment path for all of it.
Quick start
Pick how you want to run Faheem Code and start your first conversation.
Web app
The browser client: conversations, sandboxes, automations and settings.
SDK and Agent Server
Build agents in Python and drive them over REST and WebSocket APIs.
Enterprise deployment
Install on Kubernetes or a single VM, with TLS and an LLM gateway.
Where to go next
| If you want to | Read |
|---|---|
| Understand the pieces and how they fit together | Introduction |
| Write instructions the agent follows well | Prompting best practices |
| Give the agent repository context | Skills and Repository setup |
| Connect tools over MCP | Model Context Protocol |
| Run scheduled or event-driven work | Automations |
| Configure a model provider | LLM configuration |
| Self-host the backend services | Backend setup |
| Look up an endpoint | Agent Server API and Cloud REST API |
| Work out why something failed | Troubleshooting |
Open the app
Faheem Code Cloud runs the backend and sandboxes for you at app.faheemcode.ai. Everything documented here applies to it and to self-hosted installations, except where a page says otherwise.