Runs coding agents on an async-first Go harness with persisted, forkable sessions, as a library or the unreal-agent-runner CLI, for developers building or benchmarking agents.
Use it now
Keep it with up to 9 other tools you're setting up, then share them all as one link.
Try this first
Ask your coding agent:
- Agent Runner CLI
Use Unreal Agent to run an agent from a prompt or JSON request: ‹what you need›
- Harbor Benchmark Adapter
Use Unreal Agent to evaluate unreal-agent-runner: ‹what you need›
- Multi-Provider LLM Support
Use Unreal Agent to switch the LLM backend: ‹what you need›
Install the CLI
Install it as its own instructions say.
Its page: github.com/unreallabsai/unreal-agent
Call the API
Auth None needed
Its page: github.com/unreallabsai/unreal-agent
Compare
Press Compare to add it to the tray at the bottom of the window, then pick up to 2 others anywhere on the site.
- Access
- APICLI
- Auth
- None
- Skills
- 4
Skills
Agent Runner CLI
Runs an agent from a prompt or JSON request in a chosen workspace, writing events to stdout as JSONL and exiting when the task finishes.
Multi-Provider LLM Support
Switches the LLM backend between openai, openai-codex, openrouter, fireworks, and ollama with environment variables for provider and model.
Forkable Sessions
Persists append-only session history to disk so runs can be resumed or forked, with versioned, serializable session items.
Harbor Benchmark Adapter
Evaluates unreal-agent-runner with Harbor on tasks such as Terminal-Bench on Docker or Modal, producing ATIF trajectories and logs.
