BigHugger
GH Repository · deer-flow

llm-space

A desktop app to prototype agent ideas, inspect every harness step, replay failures, and evaluate performance, all in one place. Local-first, cloud-ready for managed agents.

stars
1,901
30-day movement
+289/day
Related entries
60
Connections
3
typescriptagent-appnode/bunllmmanaged-agentsharnessdevelopTypeScriptelectrobundeer-flownodellm-toolsagent

llm-space is a desktop app for prototyping agent ideas, letting you inspect every harness step, replay failures, and evaluate performance in one place. It is local-first and described as cloud-ready for managed agents.

Reach for it when you want a single desktop environment to debug, replay, and evaluate agent harness runs instead of stitching together separate tools.

Use it to

  • Prototype new agent ideas locally
  • Inspect individual harness steps during development
  • Replay failed agent runs to diagnose issues
  • Evaluate agent performance across runs
  • Work with managed agents in a cloud-ready setup

For Developers building and debugging LLM agents

Role
agent-app
Language
TypeScript
Licence
MIT
Forks
209
Open issues
5
Last push
2026-09-13
Latest release
updates · 2026-07-09
topicsagentharnessllmllm-toolsdesktop-appevaluation