DevPilot is an open-source development agent harness and orchestration platform maintained by George Eastham on GitHub. It provides a modular environment designed to execute and benchmark AI coding agent workflows across local codebases. DevPilot is structured as a TypeScript monorepo that provides both command-line execution and web-based interaction. It interacts with LLMs via API keys configured through local environment variables, processing coding tasks, file tree inspections, and multi-step agent actions directly on the developer's system. Unlike standalone single-binary coding tools, DevPilot is built around a workspace architecture that separates core agent logic, a Next.js web interface, and evaluation benchmarks. This design allows developers to test agent behavior under controlled benchmark suites while maintaining an active CLI for day-to-day coding. DevPilot is open-source software distributed via GitHub and is free to clone and run self-hosted. As of the 0.5.x release series, it is actively developed and tailored for engineers experimenting with agent orchestration.
Terminal CLI Execution — Runs agentic coding loops directly within the terminal shell. - Monorepo Architecture — Built with pnpm workspaces separating CLI packages, core libraries, and web frontends. - Built-in Benchmark Suite — Includes a dedicated benchmark harness to evaluate agent task completion and accuracy. - Next.js Companion App — Contains a web UI in apps/web for visual inspection and workflow tracking. - Environment-Driven Configuration — Uses standard .env profiles for managing API keys and LLM provider credentials.
Use cases
Use Case: Terminal-Driven Code Automation Developers use DevPilot's CLI to issue multi-step coding prompts directly from the project directory, letting the agent analyze files and make coordinated edits.
Use Case: Agent Performance Benchmarking Tooling engineers run DevPilot's benchmark suite to systematically measure model performance, prompt reliability, and tool-use fidelity against established test cases.
Use Case: Custom Agent Harness Experimentation Developers building custom developer tools can fork or extend the monorepo packages to prototype agent workflows combining terminal operations and browser interfaces.
Who it is for
DevPilot is intended for TypeScript and JavaScript engineers, AI tooling researchers, and developers who want a self-hosted, customizable agent harness. It is not designed for non-technical users or organizations seeking an enterprise-managed SaaS with zero local setup.
Install guide
Installation DevPilot is distributed as a source repository. Clone and install dependencies using pnpm: bash git clone https://github.com/geastham/devpilot.git cd devpilot pnpm install
First Run Configure your environment variables with the necessary LLM API keys: bash cp .env.example .env # Edit .env to add your model provider credentials (e.g., Anthropic, OpenAI) pnpm build
Example Session Launch the development harness using the included startup script: bash ./dev.sh
SWE-agent — SWE-agent is an open-source autonomous software engineering agent harness built and maintained by researchers from Princeton University and Stanford University.…
Agent Runbook — Agent Runbook is an open-source contract-based multi-agent skill framework developed by KnoxOps. It acts as an orchestrator and compiler that…
Weaver — Weaver is an open-source agent harness and orchestration framework maintained by sean35mm on GitHub. Built on top of the Bun…
Context Bridge — Context Bridge is an open-source CLI agent harness and context orchestration utility developed by serdardb. It is built to facilitate…
Hivelore — Hivelore is an open-source agent harness and contextual memory orchestrator created by Doucs91. It provides persistent project memory, decision tracking,…
Agentic Engineering Framework — Agentic Engineering Framework is an open-source orchestration harness and governance layer developed by Dimitri Geelen. It is designed to constrain,…