eggent-ai/eggent

An AI employee with their own computer.

Eggent – Local‑first AI workspace

What it is – Eggent is a browser‑based, self‑hosted application that lets you run LLM‑backed “agents” tied to individual projects. Each project has its own prompt, files, memory, skills, model settings and can be orchestrated together via pipelines. The UI stores everything under a local data/ folder, so you keep full control of your data.

Key features

  • Project‑scoped agents – separate instructions, files, memory and model overrides per project.
  • Local‑first storage – all data lives in data/ and can be backed up like any other files.
  • Chat UI with context – upload files or paste screenshots; the agent sees them as readable paths.
  • Pipelines – chain several project agents, passing artifacts from one step to the next.
  • External HTTP API – send a message, keep session state (project, chat, path) and receive a final reply. Useful for bots, CRMs, back‑ends, etc.
  • Telegram integration – a bot can forward messages (including voice) to Eggent; voice is transcribed locally with whisper.cpp.
  • Local voice transcription – no external speech‑to‑text service; runs on‑device via whisper.cpp.
  • Docker ready – official image on GitHub Container Registry, simple docker compose deployment, optional .env overrides.
  • Model & skill extensibility – per‑project model JSON, MCP config, and a folder‑based skill system.

How to get started

  1. Install with a single curl command (curl -fsSL https://get.eggent.ai | bash). It clones the repo, builds a Docker image and starts the container.
  2. Open http://127.0.0.1:3000, log in, and add your LLM provider keys via the UI (no need to edit .env).
  3. Create a project, upload files, and start chatting with the agent.
  4. Optionally enable the Telegram bot by providing TELEGRAM_BOT_TOKEN and webhook settings.
  5. For programmatic use, generate an External API token in Settings → API and call POST /api/external/message with a sessionId and message.

Typical use cases

  • Customer‑support assistant that remembers past tickets (project memory).
  • Research‑to‑write workflow where a “research” agent gathers sources and a “writer” agent drafts a report via a pipeline.
  • Internal tooling: connect Eggent to a CRM or ticket system through the External API.
  • Personal knowledge base: store notes, PDFs, and ask the agent to summarise them.

Limitations / notes

  • The External API returns only the final reply; streaming responses are not supported.
  • Scheduled jobs run only while the Eggent process is alive – they are not persisted across restarts yet.
  • File uploads larger than 100 MB are rejected; large folders are split into multiple requests.
  • By default the server binds to 127.0.0.1; expose it behind a reverse proxy for production.

Who might benefit – developers or teams who want a self‑hosted, project‑oriented LLM workspace with easy UI, file handling, and integration points, without relying on cloud‑only SaaS solutions.

Related

  • Project
    hAcKlyc/MyAgents

    MyAgents is an open‑source, local‑first desktop workbench for creating and orchestrating AI agents. It supports multiple LLM runtimes, a task scheduler, IM channel integration, and a private device‑to‑device network (AgentNet) for multi‑computer workflows, all under an AGPL‑v3 license.

  • Project
    NxcoreAI/EverRoom

    A local-first personal context workspace that organizes files, conversations, and memory into scoped "Context Rooms" to provide AI agents with grounded, traceable context.

  • Project
    EXXETA/exxperts

    A local-first AI agent framework that uses approval-gated memory to create persistent AI colleagues that learn and remember only what the user explicitly approves.

  • Project
    AVIDS2/memorix

    Memorix is a Node‑JS CLI/library that adds a local, SQLite‑backed, searchable memory store to any Git project. It lets AI coding assistants (Claude Code, Copilot, Cursor, etc.) share project facts, code snapshots, design rationales and commit‑derived knowledge across sessions and tools via the Model‑Context‑Protocol. Features include autopilot worksets, observation/reasoning memory, Git‑memory ingestion, code‑graph indexing, orchestration of multi‑agent tasks, a web dashboard, and optional sync. Install with npm, run `memorix setup --agent <agent>` and the agents can start reusing the shared memory.

  • Project
    nightly-labs/openbot

    OpenBot is a desktop app that lets you run and manage multiple AI agents (ChatGPT, Claude, Gemini, etc.) locally, giving each its own workspace, persistent chat, file handling, and inter‑agent messaging, with optional browser automation and a privacy‑first design.