// Local-first coding agent

A local-first coding agent that actually gets things done.

Bring your own model — an embedded GGUF file, any Ollama / LM Studio / vLLM server, or Claude, OpenAI, and Gemini with your own API keys — and get a real agent loop with a permission engine, checkpoints, and MCP tools. Nothing has to leave your machine unless you tell it to.

Download for Mac / Windows View on GitHub
Free to download No telemetry, ever Embedded models run fully offline No account required
// Why local-first

Not just another model runner

Most local-model tools — Ollama, LM Studio — stop at serving a model. localagent is a full coding agent built on top of one.

A real agent loop

Read → act → check, a deterministic permission engine (not LLM judgment), one-checkpoint-per-task revert, and per-hunk diff review before anything gets written.

Zero backend, by construction

Inference for the embedded and custom Hugging Face models runs entirely in-process — zero backend, zero telemetry. Not a setting you have to hunt down and switch off.

Rules you can read

The permission engine is plain code with four modes and a command-risk classifier, so what runs without asking is a rule you can read, not a model's guess.

PLAN DEFAULT ACCEPT_EDITS AUTO_SAFE SAFE_READ · NETWORK · DESTRUCTIVE · UNKNOWN 
# no server, no other app — a GGUF model runs in-process
node dist/cli.js "explain how add() works in math.js" \
  --workspace ./fixture-repo \
  --mode DEFAULT
// Full demo

See it end to end

Settings, MCP servers, the model picker, a custom Hugging Face search, picking a workspace, multiple session tabs, a real task with a diff to approve, and signing in — cut together from unscripted runs of the real app.

// Features

Everything it actually does

A real colored diff waiting for approval before localagent writes binary_search.py
Code safety

Per-hunk diff approval

See the real colored diff before approving an edit — not a model's summary of what it plans to do, the actual patch. Uncheck the hunks you don't want; only the checked ones get written to disk.

Searching Hugging Face's GGUF listings from inside localagent's model picker
Model registry

Search Hugging Face without leaving the app

Beyond the 10 curated models, search Hugging Face's GGUF listings from inside localagent, or type any hf:org/repo:quant path directly — same in-process runtime as the curated models, nothing downloaded until you pick one.

Adding a GitHub MCP server in localagent's MCP Servers panel
Integrations

Bring your own MCP servers

Connect local MCP servers — GitHub, Postgres, anything speaking the protocol. Every call still asks for your approval, exactly like a shell command; nothing runs silently just because it came from an MCP tool.

More under the hood

Four session tabs open at once in localagent, each keeping its own history

Multi-session tabs

Run up to 6 sessions at once. Each keeps running and updating in the background whether you're looking at it or not.

Checkpoints & revert

Every task's first write snapshots your workspace via a scratch git index — never touching your real index, HEAD, or branch. Revert restores everything, including files it created.

Command palette

Ctrl+K / ⌘K to jump to any saved session or open a panel, instead of hunting through the sidebar.

Auto-updates

A new version downloads in the background and offers a one-click restart — no manual reinstall.

The native file picker for attaching a file to a task in localagent

File attachments

Attach text and image files to a task, up to 5 at a time — stays local unless your chosen provider is a cloud model.

Claude, OpenAI, or Gemini — with automatic fallback

Add your own API key for any of the three. If the one you're using hits a rate limit or quota error mid-task, localagent automatically retries on another provider you've configured, in the order you added them — no manual switching, no lost progress.

// Compare

How it stacks up

Cursor, Copilot, and Cline are all solid tools — here's specifically where localagent differs, not a case for why the others are bad.

  localagent Cursor GitHub Copilot Cline
Runs with zero extra setup — no local server to install Yes — embedded model included — — (CLI needs Ollama/vLLM configured) — (needs Ollama configured)
Free, no subscription Yes No — from $20/mo No — subscription Yes
Bring your own API key Yes Yes (add-on to a paid plan) Yes (CLI only) Yes — the default
Deterministic permission engine (not model judgment) Yes — 4 modes, command-risk classifier Partial Partial Yes — per-action approval
Checkpoints & one-click revert Yes Partial Partial Yes
Multiple sessions running at once Yes — up to 6 tabs Limited Limited Single session
Open source No — source-viewable, proprietary No No Yes — Apache-2.0

Accurate as of October 2026, based on each tool's public pricing pages and docs — pricing and features change; if something here is outdated, let us know.

// Download

Free, and complete

Every embedded model, the custom-server option, and your own Claude, OpenAI, or Gemini API key — no feature is gated behind a paywall.

Download for MacApple Silicon · .dmg Download for Windowsx64 · installer .exe

Beta builds are unsigned. Your OS shows a one-time warning on first launch — normal for a beta, not a sign anything's wrong. On macOS: right-click the app → Open → confirm. On Windows: "More info" → "Run anyway." Only needed once.

Support the project

A one-time payment that unlocks no feature — the app is already free and complete either way. It's just a way to say thanks if localagent has been useful to you.

Coming soon
// FAQ

Questions? We've got answers.

Is localagent really free?

Yes. Every embedded model, custom Hugging Face models, the custom-server option, and your own Claude, OpenAI, or Gemini API key are all included — nothing is gated behind a paywall. The optional one-time "Support the project" payment unlocks no feature; it's just a way to say thanks.

Does my code ever leave my machine?

Not by default. The embedded and custom Hugging Face models run entirely in-process — zero backend, zero telemetry. Your code only leaves your machine if you explicitly choose a cloud provider — Claude, OpenAI, or Gemini, each sending task context to that provider's own API with your own key — or point it at a custom OpenAI-compatible server.

What platforms does it support?

macOS (Apple Silicon) and Windows (x64) today, as prebuilt installers. Builds are currently unsigned, so your OS shows a one-time warning on first launch.

Is localagent open source?

No. The source is publicly viewable on GitHub, but it's proprietary — all rights reserved, no license is granted to copy, modify, or redistribute it. The prebuilt installers are free to use as distributed.

Do I need an account to use it?

No. Signing in with Google is entirely optional and only used to back up session history to your own Drive — it gates no feature.

For support, please open an issue on GitHub.

// Legal

Privacy Policy

Last reviewed: September 2026. Written to accurately describe what the app does — not a substitute for professional legal advice. Treat it as an honest draft, not a finished legal document.

There is no backend for this project — no server we operate, no analytics, no telemetry, nothing phoning home. Everything below happens directly between your machine and whichever service you've explicitly configured.

Your code and files stay on your machine unless you choose a provider that sends them elsewhere: Claude, OpenAI, and Gemini each send task context to that provider's own API, using your own API key — nothing is sent to us. A custom server sends it to whatever OpenAI-compatible endpoint you point at. If you've configured more than one cloud provider and the active one hits a rate limit or quota error mid-task, the same task's context is automatically sent to the next configured provider instead — never to one you haven't added a key for. The default embedded mode — curated or a custom Hugging Face model — sends nothing anywhere once downloaded.

Session history is saved locally on your machine. It only leaves your machine if you sign in with Google, in which case it's backed up to a hidden, app-private folder in your own Google Drive — not visible in your normal Drive UI, not accessible to any other app or person.

Google sign-in is optional and gates nothing. If used, your email, name, and profile picture URL are requested from Google's own identity endpoint and stored locally; the refresh token is encrypted at rest via your OS's native credential store, never written to disk as plain text.

Crash and error logs are written locally and never uploaded anywhere — no external crash-reporting service, no account, no telemetry.

The optional "Support the project" purchase is processed entirely by the payment platform used to sell it — we do not receive or store your payment details ourselves. It unlocks no feature and creates no account.

Questions about this policy: lavuchandu169@gmail.com.

// Legal

Terms of Service

Last reviewed: September 2026. Written to accurately describe the terms this project intends — not a substitute for professional legal advice.

localagent is provided as-is, with no warranty of any kind, express or implied. You use it at your own risk, particularly in modes that let it edit files or run shell commands without per-action approval.

License. The source code is publicly viewable but proprietary — all rights reserved. No license to use, copy, modify, or redistribute the source is granted. The prebuilt installers are provided for personal use of the app as distributed.

The optional "Support the project" purchase does not create a contract for any feature, service, or ongoing obligation beyond the payment itself. Because it unlocks nothing and the app remains fully functional either way, all such purchases are final and non-refundable.

To the maximum extent permitted by law, the author is not liable for any damages arising from use of this software, including data loss or unintended file or system changes made by the agent while running with your permission.

Questions about these terms: lavuchandu169@gmail.com.