Skip to content
/ rapid Public

Multi-agent development orchestration system with Claude, OpenCode, Aider and more

Notifications You must be signed in to change notification settings

a3tai/rapid

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RAPID Development Framework

Research. Augment. Plan. Integrate. Develop.

RAPID orchestrates AI coding assistants within containerized development environments. It wraps tools like Claude Code, OpenCode, and Aider to provide a unified, secure, and reproducible developer experience.

What RAPID Does

  • Manages dev container lifecycle - rapid start / rapid stop handle everything
  • Auto-configures AI coding tools - Installs and configures Claude, OpenCode, Aider inside containers
  • Handles secrets injection - 1Password, Vault, or environment-based secret loading
  • Generates agent instruction files - Creates AGENTS.md, CLAUDE.md with project context
  • Supports concurrent AI agents - Run multiple agents in tmux-style panes

Quick Start

npm install -g @rapid-dev/cli
rapid init
rapid start
rapid dev

Commands

Command Description
rapid init Initialize RAPID in current project
rapid start Start container, load secrets, prepare environment
rapid dev Launch AI coding session
rapid dev --multi Launch multiple agents concurrently
rapid stop Stop container and cleanup
rapid agent <name> Switch or add AI agent
rapid mcp Manage MCP servers (add/remove/list/status)
rapid secrets Manage project secrets
rapid config View/edit configuration

Supported AI Tools

Tool CLI Status
Claude Code claude Supported
OpenCode opencode Supported
Aider aider Supported
Cursor Editor mode Planned
GitHub Copilot CLI gh copilot Planned

The RAPID Methodology

RAPID is both a tool and a methodology for effective AI-assisted development:

Phase Purpose
Research Gather context before engaging AI - codebase structure, docs, patterns
Augment Enhance with external knowledge - APIs, documentation, MCP servers
Plan Structure work before execution - task breakdown, acceptance criteria
Integrate Ensure environment readiness - containers, secrets, tooling
Develop Execute with AI assistance - generate, test, iterate, review

MCP Servers

RAPID supports Model Context Protocol (MCP) servers to extend AI agent capabilities:

# List available MCP server templates
rapid mcp list --templates

# Add servers
rapid mcp add context7
rapid mcp add tavily
rapid mcp add playwright

# Check status
rapid mcp status

Built-in templates include:

  • context7 - Documentation context for libraries
  • tavily - Web search and data extraction
  • playwright - Browser automation
  • github - GitHub operations
  • filesystem - File system access
  • memory - Persistent knowledge graph

Configuration

RAPID uses rapid.json for project configuration:

{
  "version": "1.0",
  "agents": {
    "default": "claude",
    "available": {
      "claude": {
        "cli": "claude",
        "instructionFile": "CLAUDE.md"
      }
    }
  },
  "secrets": {
    "provider": "1password",
    "vault": "Development"
  }
}

Documentation

Concepts

Guides

Reference

Examples

Architecture

flowchart TB
    subgraph CLI["RAPID CLI"]
        commands["rapid init | rapid start | rapid dev | rapid stop"]
    end

    subgraph Container["Dev Container"]
        subgraph Agents["AI Agents"]
            claude["Claude Code"]
            opencode["OpenCode"]
            aider["Aider"]
        end
        resources["AGENTS.md | CLAUDE.md | Secrets | MCP Servers"]
    end

    subgraph Codebase["Your Codebase"]
        files["Source Files"]
    end

    CLI --> Container
    Container --> Codebase
Loading

Requirements

  • Docker Desktop, Podman, or compatible container runtime
  • Node.js 18+
  • API key for your AI provider (Anthropic, OpenAI, etc.)

License

MIT

About

Multi-agent development orchestration system with Claude, OpenCode, Aider and more

Resources

Stars

Watchers

Forks

Packages

No packages published