> ## Documentation Index
> Fetch the complete documentation index at: https://docs-dev.byterover.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# OpenClaw Integration

> Configure ByteRover as long-term memory for OpenClaw agents

<img src="https://mintcdn.com/demo-a750c123/yDOWPuEFbFlK9TsA/images/byterover_x_openclaw.png?fit=max&auto=format&n=yDOWPuEFbFlK9TsA&q=85&s=ed0c33fbd01aa11729bffffdb3083710" alt="ByteRover x OpenClaw" style={{ borderRadius: '12px', width: '100%' }} width="2100" height="700" data-path="images/byterover_x_openclaw.png" />

Give your [OpenClaw](https://docs.openclaw.ai) agents persistent memory with ByteRover. One setup script installs context retrieval and automatic memory curation — so your agents remember decisions, patterns, and fixes across sessions.

## Prerequisites

Ensure the following are installed and available in your `PATH`:

| Requirement                   | How to install                                                        |
| ----------------------------- | --------------------------------------------------------------------- |
| **Node.js**                   | [nodejs.org](https://nodejs.org)                                      |
| **OpenClaw CLI** (`openclaw`) | [docs.openclaw.ai/install](https://docs.openclaw.ai/install#npm-pnpm) |

The script verifies each prerequisite and stops with a clear error message if anything is missing. If the `brv` CLI is missing, the script installs it from `https://byterover.dev/install.sh`. The `byterover` Clawhub skill is installed automatically via `npx clawhub@latest install byterover --force`.

## Quick Start

```bash theme={null}
curl -fsSL https://byterover.dev/openclaw-setup.sh | sh
```

The interactive script will:

1. Verify prerequisites and back up your config
2. Prompt you to enable each feature (Context Engine, Memory Flush)
3. Update your workspace protocol files (`AGENTS.md`, `TOOLS.md`) with ByteRover commands

After it finishes, the script prints "Connect an LLM provider" as the next step — see [Next Steps](#next-steps) below.

For a minimal integration that only installs the ByteRover Skill (no plugin), use the [connector install](/connectors/overview#connect-an-agent) instead.

## Features

The setup script configures two optional features. You choose which to enable during the interactive setup.

### Context Engine

Automatically retrieves relevant knowledge before each agent response and stores new insights after each turn. Installs the [`@byterover/byterover`](https://www.npmjs.com/package/@byterover/byterover) plugin as an OpenClaw [context engine](https://docs.openclaw.ai/concepts/context-engine). Requires OpenClaw v2026.3.22+.

You can also install directly with `openclaw plugins install @byterover/byterover` — see the [full setup guide](https://www.npmjs.com/package/@byterover/byterover) for details.

### Automatic Memory Flush

<Warning>
  Experimental — behavior may change in future updates.
</Warning>

Automatically curates insights to ByteRover when the agent's context window approaches compaction. The agent reviews the session for architectural decisions, bug fixes, and patterns, then runs `brv curate` to save them before context is compacted. Learn more about [compaction behavior](https://docs.openclaw.ai/concepts/memory#automatic-memory-flush-pre-compaction-ping).

## Next Steps

### Connect an LLM provider

ByteRover CLI needs an LLM provider to enable its agent features.
Run one of the following to get started:

```bash theme={null}
brv providers connect byterover         # Free, requires login with ByteRover account
brv providers connect <llm-provider>    # Requires 3rd API key: OpenAI, Anthropic ...
```

Supported list: OpenRouter, Anthropic, OpenAI, Google Gemini, xAI, Groq, Mistral, DeepInfra, Cohere, Together AI, Perplexity, Cerebras, Vercel, MiniMax, GLM, Moonshot AI, and OpenAI Compatible.

To see all available providers: `brv providers list`

<CardGroup cols={2}>
  <Card title="Reference" icon="book" href="/autonomous-agents/openclaw-reference">
    Configuration details, troubleshooting, and advanced topics
  </Card>
</CardGroup>