DeepRise is an autonomous multi-agent AI coding system that builds, tests, and improves apps end-to-end with minimal input. Once given a direction, swarms of agents continue working independently until the task is done — including proper real-user-style testing and iteration.
- Super Agent: Acts as the CEO of the agent swarm — coordinates all other agents, receives and consolidates their feedback, sets strategies, and instructs each agent on what to do to drive the project forward.
- Orchestration Agent: Designs software architecture, creates implementation plans, and generates detailed TODO lists for developers.
- Developer Agent: Implements features, writes code, and coordinates with the test agent to ensure functionality.
- Test Agent: Creates and runs comprehensive tests to validate code quality and reliability.
- User-Level Test Agent: Simulates real-world user interactions to identify issues like a real human user.
- Runtime QA Agent: Runs the application in an isolated environment, validates real user flows, and coordinates fixes with other agents.
- Explore Agent: Searches and analyzes codebases to answer questions or locate specific patterns.
- Code Review Agent: Reviews code for adherence to best practices, readability, and maintainability.
- Security Agent: Identifies and mitigates security vulnerabilities in the code.
- DevOps Agent: Automates deployment pipelines, monitors infrastructure, pushes to github, and ensures smooth operations.
DeepRise uses a Retrieval-Augmented Generation (RAG) approach to manage memory efficiently, ensuring relevant context is always available.
git clone https://github.com/MehdiGhorb/deeprise.git
cd deepriseTwo installation methods:
-
Install from Remote Repository:
bash install
-
Install from Local Repository:
bash install --local
This project was inspired by and built upon the foundations of OpenCode. Special thanks to the OpenCode community for their contributions and innovation.