Skip to main content
ByteRover CLI gives your AI coding agents persistent memory — stored locally on your machine, no account required. Connect your own LLM (BYOK), install your preferred coding agents, and optionally sync to the cloud when you’re ready to collaborate.

​
Get started in 2 minutes

1

Install ByteRover CLI

macOS Apple Silicon (M1/M2/M3+), Linux x86_64, and Linux ARM64 only. On macOS Intel or Windows — use the npm tab.
curl -fsSL https://byterover.dev/install.sh | sh
Restart your shell after install, or run the export PATH line the installer prints.
2

Open your project

Open your project in a terminal or in your IDE with a coding agent, then run brv — no sign-up required:
cd path/to/your/project
brv
3

Configure and connect

The built-in LLM is free with limited credits — no API key required. Or connect any of the 18 supported external providers:
Connect your LLM provider (OpenRouter, Anthropic, OpenAI, and more):
/providers
Select a model from your connected provider:
/model
Connect your coding agent (Cursor, Claude Code, Windsurf, and more):
/connectors

Continue with the Quickstart

Full setup guide: install options, provider configuration, and your first context

​
Core concepts

Context Tree

Your project’s knowledge base — stored as markdown files in .brv/context-tree/. Human-readable, git-friendly, and hierarchically organized by domain and topic.

LLM Providers

Connect 18 external providers (OpenRouter, Anthropic, OpenAI, and more), run a local model, or use the built-in LLM for free. Bring your own API key.

Daemon-First Architecture

A background daemon handles all execution. CLI commands and coding agents connect directly — no TUI required. Auto-starts on demand, no manual setup.

BRV Hub

Community marketplace for pre-built agent skills and context bundles. Install in one command and extend what your agents can do.

Local vs Cloud

Everything works offline by default. Cloud adds team sync, multi-machine access, and backup — without changing your local workflow.