humanlayer/humanlayer

Project overview

It addresses the concrete pain of running and steering multiple AI coding agents across machines and sessions, unifying tasks, plans, and reviews with plugins that mirror Claude Code, OpenCode, and Pi sessions.

Project type

AI Agent · AI Coding

Use cases

Coding & Development

Deployment

Refer to project documentation

License

Apache-2.0

Best for

• Developer teams that need to run and steer multiple AI coding agents across machines and sessions from one place.
• Teams already using Claude Code, OpenCode, or Pi who want their sessions mirrored into a shared workspace.

Key capabilities

• Runs coding agents (Claude Code, Codex, OpenCode, Pi) on your machines and lets your team steer them from one place: tasks, sessions, plans, and reviews.
• Plugins mirror Claude Code, OpenCode, and Pi sessions into HumanLayer, plus built-in HumanLayer skills and Research-Plan-Implement skills.

Limitations and risks

• Guides for running HumanLayer yourself are in progress, so expect to rely on the source-based build path when self-hosting.

Getting started

• Setup is demanding: building the monorepo requires Bun and Docker, and the native desktop app additionally requires Rust. First steps: install Bun and Docker, run bun install, copy .env.example to .env, then run bun --bun run devstack help to run the whole stack locally.

Evidence and sources

• README: HumanLayer runs coding agents (Claude Code, Codex, OpenCode, Pi) on your machines and lets your team steer them from one place: tasks, sessions, plans, and reviews, synced between…
• README: - `apps/riptide-claude-code-plugin`, `apps/riptide-opencode-plugin`, `apps/riptide-pi-extension`: plugins that mirror Claude Code, OpenCode, and Pi sessions into HumanLayer
• README: You need [Bun](https://bun.sh) and Docker. Building `apps/riptide-native` also needs Rust.
• README: HumanLayer Cloud runs this code. Guides for running it yourself are in progress.

Source updated at: 2026-10-23T18:25:49Z

Open original GitHub project