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.

What it solves

LLM Space is a desktop application designed for agent builders to prototype, debug, and evaluate AI agents in a single environment. It eliminates the need to switch between different tools for prompt engineering, execution tracing, and performance measurement.

How it works

Built as a TypeScript monorepo using Bun and Electrobun, the app provides a local runtime for managing agents. It integrates with the Pi Agent Core framework to allow users to build agents, manage prompts and tools, and track execution flows. The app stores API keys and thread data locally on the user's machine for privacy.

Who it’s for

It is primarily for developers and AI engineers who are building agentic workflows and need a visual environment to inspect model calls, debug failures through history replay, and iterate on agent logic.

Highlights

  • Integrated Build & Trace: Write prompts and model settings while seeing every model call and tool run in real-time.
  • Execution Debugging: Replay historical runs and step through them to identify specific failure points.
  • Agent Generation: Uses AI to help write prompts and tools, and can export threads into runnable LangGraph agents.
  • Local-First Management: Keeps all threads, files, and API keys stored locally on the user's computer.

Related

  • Project
    hAcKlyc/MyAgents

    An open-source desktop workspace for personal AI agents that integrates local files, terminals, and a task scheduling system to move beyond simple chat interfaces.

  • Project
    proma-ai/Proma

    Proma is an open‑source, cross‑platform desktop AI agent that combines chat, planning, project‑based memory, and embedded tools (browser, terminal, calendar, Obsidian‑style markdown editor). It supports parallel sub‑sessions, skill bundles, and remote connectors (Slack, WeChat, etc.). Installable via binaries for macOS, Windows, and Linux; the open‑source edition requires user‑provided LLM APIs, while a commercial edition adds hosted model access, team management, and extra services. Licensed under AGPL‑3.0 with optional commercial licensing.

  • Project
    stevibe/BenchLocal

    A local-first desktop application for running and comparing LLM performance across various benchmark sets called Bench Packs against local or remote models.

  • Project
    EKKOLearnAI/hermes-studio

    A local-first AI workspace that consolidates multi-agent chat, coding, and visual workflows into a single desktop or web console.

  • Project
    simstudioai/sim

    A workspace to build, deploy, and manage AI agents and workflows, featuring a visual builder, 1,000+ integrations, and integrated data management.