This extension provides a Model Context Server for ByteRover, for use with the Zed AI assistant.

ByteRover is a self-improving, shared memory layer that remembers your AI agent's coding experiences—enabling smarter, faster development and seamless collaboration across your entire dev team. Learn more about ByteRover.

To use the extension:

1. Visit ByteRover to create a memory workspace
2. Follow the Space start guide to get your configuration values
3. Configure the context server by setting the auth_header in your Zed <project-root>/.zed/settings.json

{
  "context_servers": {
    "mcp-server-byterover": {
      "source": "extension",
      "settings": {
        "auth_header": "YOUR_SPACE_AUTH_HEADER"
      }
    }
  }
}

The ByteRover MCP server provides two core memory management tools:

• byterover-retrieve-knowledge: Search your curated knowledge base of past coding solutions and insights to find relevant help for your current task
• byterover-store-knowledge: Capture key coding insights from human-agent interactions to help solve similar tasks in the future

The ByteRover extension uses a flexible rule system that allows you to customize how the AI agent interacts with your memory workspace. Rules define when and how the agent should use memory operations during development workflows.

# Important: Automated memory workflow
always use byterover-retrieve-knowledge tool to get the related context before any tasks
always use byterover-store-knowledge to store all the critical informations after successful tasks

This configuration ensures the agent:

1. Searches existing memories before starting any development task to leverage past solutions
2. Stores new insights after successfully completing tasks to build your knowledge base

You can configure different triggers for memory operations:

# Search memory before specific tasks
always use byterover-retrieve-knowledge before implementing new features
always use byterover-retrieve-knowledge when debugging complex issues
always use byterover-retrieve-knowledge for architecture decisions

# Store memories based on task completion
always use byterover-store-knowledge after successful bug fixes
always use byterover-store-knowledge when completing code reviews
always use byterover-store-knowledge for performance optimizations

Create a <project-root>/.rules file and add your custom rules to control how the agent uses memory:

# Context-specific memory usage
use byterover-retrieve-knowledge for similar technology stacks only
use byterover-store-knowledge to store reusable patterns not full implementations
focus byterover-store-knowledge on design decisions and lessons learned

# Memory format guidelines
store coding patterns and architectures in byterover-store-knowledge
capture algorithm explanations and trade-offs in byterover-store-knowledge
include dependency recommendations and gotchas in byterover-store-knowledge

# Conditional memory usage
use byterover-retrieve-knowledge when working with unfamiliar libraries
use byterover-retrieve-knowledge before refactoring legacy code

# Team collaboration
use byterover-retrieve-knowledge to maintain coding standards across team
store byterover-store-knowledge with clear problem-solution mapping

Beyond automated rules, you can explicitly request memory operations:

• "Search our memories for React hook patterns" - Manually trigger memory search
• "Remember this API integration pattern for future use" - Manually store knowledge
• "What do we know about handling database migrations?" - Query specific domain knowledge

1. Start Simple: Use the default rules initially, then customize based on your workflow
2. Focus on Patterns: Store reusable coding patterns, not complete implementations
3. Context Matters: Include relevant tech stack and project context in memory rules
4. Iterate Rules: Refine your rules based on how effectively they improve your development speed

Add your custom rules to your workspace configuration in one of these locations:

• .zed/settings.json (project-specific)
• Global Zed settings (applies to all projects)

For more information about ByteRover's features, pricing, and capabilities, visit the ByteRover homepage.

Explore how ByteRover can 10x your development efficiency with agentic memory management and team collaboration features.