Open-source long-running coding agents that work in parallel
DeepRise is a multi-agent system where autonomous agents plan, build, test, and improve software in parallel.
curl -fsSL https://deeprise.dev/install | bashOpen-source multi-agent system for autonomous coding
DeepRise is a multi-agent system where long-running autonomous coding agents work in parallel to plan, build, test, and improve software. Each agent in the swarm has a specialized role — from architecture design and feature implementation to automated testing, code review, security analysis, and deployment.
Unlike single-prompt coding assistants, DeepRise agents persist across multi-step workflows. A Super Agent coordinates the swarm, delegates tasks to specialized agents, and consolidates feedback until project goals are met. This makes DeepRise suited for long-running development tasks that span many files, many tests, and many iterations.
DeepRise is open source under the MIT license. Install it locally, inspect the agent orchestration code, customize agent behavior, and contribute on GitHub.
Agents
A swarm of specialized autonomous agents — orchestration, development, testing, review, security, and DevOps — working in parallel on long-running tasks
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.