Skip to content

Repository files navigation

BMAD CrewAI Integration

CrewAI orchestration layer for the BMAD-Method framework

This package provides a structured integration between CrewAI's multi-agent orchestration capabilities and the BMAD-Method framework. CrewAI acts as the primary orchestrator that reads BMAD templates and coordinates specialized BMAD agents to write artefacts to BMAD-specified folder structures.

Architecture Overview

  • CrewAI: Primary orchestration engine that manages workflow execution
  • BMAD Templates: YAML templates defining agent roles, workflows, and artefact structures
  • BMAD Agents: Specialized agents (PM, Architect, QA, Dev, PO, SM) with domain expertise
  • Artefact Output: All outputs written to BMAD folder structure (docs/, stories/, qa/, etc.)

Key Features

  • Template-Driven Orchestration: CrewAI reads BMAD YAML templates to understand workflow requirements
  • Agent Coordination: Seamless coordination between BMAD specialized agents
  • Quality Assurance: Built-in quality gates and validation checkpoints
  • File System Integration: Artefacts automatically written to BMAD folder conventions
  • Process Integrity: Maintains BMAD methodology execution flow
  • Simple Credential Storage: Environment variables with JSON config fallback
  • Basic API Rate Limiting: Simple request spacing with retry on rate limits
  • Essential Error Handling: HTTP status checking with basic error recovery

Quick Start (MVP)

Simple Setup

# Install
pip install -e .

# Set API keys (environment variables)
export API_OPENAI_KEY="your-openai-key"
export API_ANTHROPIC_KEY="your-anthropic-key"

# Test connection
python -m bmad_crewai.simple_cli test

Basic Usage

from bmad_crewai.simple_api import BMADCrewAISimple

async def main():
    bmad = BMADCrewAISimple()

    # Get API client
    client = bmad.get_api_client("openai")

    # Make request with basic rate limiting
    async with client:
        result = await client.make_request("GET", "https://api.openai.com/v1/models")
        print(result)

asyncio.run(main())

Security & API Management

Simple Credential Storage

  • Environment Variables: API_OPENAI_KEY, API_ANTHROPIC_KEY
  • Config File: Simple JSON storage in ~/.bmad-crewai/config.json
  • Priority: Environment variables override config file

Basic Rate Limiting & Error Handling

  • Simple Rate Limiting: 1 second delay between requests
  • Basic Error Handling: HTTP status code checking with simple retry
  • API-Specific Headers: Automatic header injection for OpenAI/Anthropic

About

BMAD integration with CrewAI agents

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages