Skip to content

Simulating an engineering crew of agent comprising of a lead engineer, frontend engineer, backend engineer,test engineer

Notifications You must be signed in to change notification settings

Gracelesslemon/Engineering-crew-of-agents

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 

Repository files navigation

Engineering Crew: AI-Powered Software Development Team

A CrewAI project that orchestrates a team of specialized AI agents to collaboratively design, implement, and test software applications.

Overview

This project demonstrates autonomous software development using CrewAI's multi-agent framework. Four specialized agents work together to transform high-level requirements into an application.

Agent Architecture

All agent and task configurations are defined as YAML files in the config/ folder for easy customization.

Engineering Lead

Translates high-level requirements into detailed technical specifications and system designs for the development team.

Backend Engineer

Implements Python modules based on the engineering lead's specifications. Executes code safely within isolated Docker containers to ensure system security.

Frontend Engineer

Builds interactive Gradio user interfaces according to design specifications, creating intuitive interfaces for end users.

Test Engineer

Develops comprehensive unit tests to validate functionality and ensure code quality.

Tech Stack

  • Package Manager: uv
  • AI Framework: CrewAI
  • LLM: Gemini 2.0
  • UI Framework: Gradio
  • Containerization: Docker

Screenshots:

Note

No directions were given stating how the ui should look.

Account Actions page:

image

Trade page:

image

Reports page:

image

Current Limitations

  • Code Formatting: Despite explicit instructions to avoid markdown formatting, Gemini 2.0 occasionally wraps code in backticks, requiring manual cleanup
  • Test Completeness: Generated unit tests sometimes miss boilerplate code like:
  if __name__ == "__main__":
      unittest.main()
  • Other minor issues exist, but they can be easily addressed with sufficient LLM API resources and prompt optimization.

About

Simulating an engineering crew of agent comprising of a lead engineer, frontend engineer, backend engineer,test engineer

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages