HermOS-IDE

mcp
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 7 GitHub stars
Code Warn
  • process.env — Environment variable access in next.config.ts
  • process.env — Environment variable access in scripts/ensure-extra-prisma-engines.mjs
  • network request — Outbound network request in scripts/ensure-extra-prisma-engines.mjs
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Local-first, open-source AI agent desktop IDE. Autonomous subagents edit code, run terminal commands, and automate browsers natively on your machine with 100% privacy and zero cloud telemetry.

README.md

HermOS IDE — Local-First AI Coding Environment

Your code stays on your machine. Your models, your keys, your data — never leaves your device.

HermOS IDE is a desktop code editor with a built-in autonomous coding agent. It runs entirely offline on your computer: edit files, run commands, browse the web, and generate documents — all powered by your own AI provider keys. No accounts, no cloud, no telemetry.

HermOS IDE


Why HermOS?

  • 100% Local & Private — All data lives in a local SQLite database on your machine. No remote servers, no tracking.
  • Bring Your Own Keys — Works with OpenAI, Anthropic Claude, Groq, Mistral, Together, Ollama (local), and more. You control which model runs.
  • Agent That Codes — Chat, Architect, and Agent modes handle everything from quick questions to multi-step refactors.
  • Works Offline — Once installed, the editor, terminal, and local models (Ollama) run without internet.
  • Extensible — Model Context Protocol (MCP) servers, plugins, and skills extend the agent with your own tools.

Features

Autonomous Agent Multi-step coding tasks: read, edit, run, test — locally.
Integrated Terminal Full shell access rooted in your project folder (commands run with your OS privileges).
Built-in Browser Headless browser panel synced to the editor for live previews.
MCP Support Connect any stdio/SSE MCP server (filesystem, GitHub, databases…).
Office Builder Generate PPTX, DOCX, and PDF directly from the agent.
Plugins & Skills Install community tools or build your own.
Cross-Platform Native installers for Windows, macOS (Apple Silicon & Intel), and Linux.

Download

Get the latest release for your platform:

Platform Download
Windows HermOS IDE for Windows — grab the .msi under Assets
macOS HermOS IDE for macOS — Apple Silicon & Intel, grab the universal .dmg
Linux HermOS IDE for Linux — grab the .deb under Assets

Or browse all assets on the Releases page.

Website & Docs: https://hermos.is-a.dev/


Getting Started (3 Steps)

1. Install

Download and run the installer for your OS. On first launch HermOS creates its local data folder and encryption keys — no setup wizard, no login.

2. Add a Provider Key

Open Settings → Providers and paste an API key:

  • Cloud: OpenAI, Anthropic, Groq, Mistral, Together, OpenRouter, NVIDIA NIM, Zen …
  • Local (offline): Ollama — run ollama serve and add http://localhost:11434/v1

Keys are encrypted at rest with a per-install secret.

3. Start Coding

  • Open a folder — click the project selector in the top bar → “Open Folder / New Project” (desktop) or use the Files panel, or drop files into the workspace.
  • Ask the agentChat for Q&A, Architect for planning, Agent for autonomous execution.
  • Run & preview — The terminal and browser panel are right inside the IDE.

Privacy & Security

  • Local-first — Database, uploads, and secrets live under your OS app-data folder (%APPDATA%\com.hermos.ide on Windows, ~/.config on Linux, ~/Library/Application Support on macOS). Nothing is sent to HermOS servers — there are none.
  • Encrypted keys — Provider API keys are stored AES-256-GCM encrypted. The encryption key is generated per install and never leaves your device.
  • No telemetry — No analytics, no crash reporting, no update phoning home beyond the optional GitHub Releases check.
  • Loopback only — The bundled server binds to 127.0.0.1 and rejects non-local requests.

System Requirements

  • Windows 10/11 (x64) · macOS 12+ (Apple Silicon & Intel) · Linux (Ubuntu 22.04+ / Debian)
  • 4 GB RAM minimum, 8 GB recommended
  • ~300 MB disk for the app + space for your projects
  • For local models: Ollama + a capable GPU recommended

FAQ

Is HermOS free?
Yes — MIT licensed. You only pay your AI provider (or run Ollama for free locally).

Does it need internet?
Only to call cloud model APIs. The editor, terminal, file tools, and Ollama work fully offline.

Where is my data?
On your machine. Uninstalling the app does not delete the data folder — remove it manually if desired.

Can I use it on a shared machine?
HermOS is single-user, bound to the OS account that installed it.


Support & Contributing


License

MIT — see LICENSE.

Reviews (0)

No results found