wecode-ai/Wegent

Plan, build, and deliver with an open-source, self-hostable AI workspace for coding, collaboration, and automation.

Wegent – an open‑source AI‑powered work system

What it is – Wegent is a suite of tools that let you use large‑language‑model (LLM) agents to help with software development across a single computer, a team, or a fleet of remote machines. It consists of three main parts:

Component Role
Wegent Desktop An Electron‑based workbench that lives on your local machine. It shows projects, tasks, chat with the coding agent (named Codex), test results, file diffs and command output all in one view.
Wegent Web A browser UI for remote agents, knowledge bases, automation rules and admin functions.
Wegent Backend Self‑hostable server that stores shared project spaces, model configurations, execution queues and permissions. It also runs the Rust‑based task executor that talks to the Codex agent.

Key features

  • Unified task view – conversations, tool activity, test runs, changed files and diffs are kept together.
  • Reusable agents – combine LLM models, prompts, “Skills” (tool wrappers), knowledge stores and collaboration settings.
  • Project spaces – shared Kanban board, files, discussions, automation pipelines and execution history for teams.
  • Multiple execution targets – run a task locally, on a remote work machine, or on a server‑managed executor.
  • Self‑hosting – deploy the Web UI and Backend on‑premises; control models, scheduling, permissions and knowledge services.

How it works

  • The desktop host is an Electron process that loads the DSH (DeepSeek Harness) runtime. DSH discovers plugins that provide UI, terminal, and executor capabilities.
  • The Executor (written in Rust) receives task commands from the UI, drives the Codex coding agent, and streams status, tool events and results back to the UI.
  • Codex reads and modifies the local project files, runs commands, and can invoke tools (e.g., a Gantt‑chart generator shown in the screenshots).
  • When connected to the Backend, the same project data, agents and execution devices are shared with Wegent Web, enabling collaboration and remote execution.

Getting started

  1. Desktop – download the installer from the releases page, add a local project, create a task, and watch the agent work.
  2. Web & Backend – run the provided install.sh script with the --standalone flag. It builds Docker images, starts the services, and exposes the UI at http://localhost:3000 where you can set an admin password and configure a model.

Development

  • Requires Node 20+ and pnpm. The workspace contains many packages (desktop, mobile, web, backend, Rust executor, knowledge services, etc.).
  • Typical commands: pnpm install, pnpm --filter wework dev for the desktop, pnpm --filter wework dev:mac for macOS, and the Rust executors are built from the executor/ and executor_manager/ directories.

Who it’s for – developers who want an integrated LLM‑assistant for coding that can be run locally or self‑hosted for a team, with the ability to route work to remote machines and keep all artefacts (code, tests, discussions) in one place.

Links

Related

  • Project
    eggent-ai/eggent

    Eggent is a self‑hosted, browser‑based AI workspace that lets you run LLM‑backed agents per project, store files and memory locally, chain agents with pipelines, and integrate via a simple HTTP API or Telegram bot. It ships as a Docker image, uses local storage under `data/`, and supports on‑device voice transcription with whisper.cpp.

  • Project
    zhayujie/CowAgent

    CowAgent is an open‑source, self‑hosted AI assistant that plans tasks, runs tools, keeps a personal knowledge base, and can operate as a team of cooperating agents. It supports many LLM providers, multimodal input, dozens of chat channels (Telegram, Slack, WeChat, etc.), a plug‑in skill system, and a three‑tier memory with nightly “Deep Dream” distillation. Install via a one‑line script or Docker, manage everything through a web console, and extend the system through the public Skill Hub or custom tool integrations.

  • Project
    eigent-ai/eigent

    An open-source desktop application for building and deploying a multi-agent AI workforce to automate complex digital workflows.

  • Project
    agnt-gg/agnt

    AGNT is a local‑first, self‑hostable desktop app that bundles AI agents, visual workflows, long‑running goals, memory, plugins and a local HTTP API into one “agent operating system”. Built with Electron, Vue, Express and SQLite, it supports 15+ LLM providers, 60+ workflow nodes, plugin templates and SkillForge for self‑improving skills. Ideal for developers or teams that need durable, inspectable AI automation on‑premise.

  • Project
    kagent-dev/kagent

    kagent is a CNCF‑incubated, open‑source framework that lets you define and run AI agents as native Kubernetes custom resources. Agents are described in YAML, can use any major LLM provider, run on Go/Python runtimes or custom back‑ends, and have built‑in tool integration, persistent sessions, and OpenTelemetry tracing. The system includes a controller, web UI, CLI, PostgreSQL storage, and a substrate that manages agent lifecycles, making AI workloads manageable with standard Kubernetes/GitOps workflows.