RAVINDRAN S.SDE
LocalMind icon
Developer ToolsActive development

LocalMind Agent SDK & Runtime

Loopback-only local AI agent runtime, terminal assistant, and Python streaming SDK.

A private, local-first LLM orchestration system that connects your terminal, local workspace index, and host telemetry through a lightweight Rust launcher and typed Python SDK without sending data to third-party clouds.

Release Versionv0.1.0
ArchitectureUniversal
PlatformsLinux
LicenseOpen Source (MIT/GPL)
localmind · console
VERIFIED
$ quick_launch
git clone --depth 1 https://github.com/ravindran-dev/localmind.git "$HOME/.local/share/localmind" && bash "$HOME/.local/share/localmind/install.sh"
Primary StackRust Native Launcher, FastAPI / Uvicorn
LanguagesPython 3.12+, Rust, TypeScript
Release StageActive development
IntegritySHA-256 Audited
Manifest Revision:0.1.0 (Native Launcher & Rust daemon)

Official Packages & File Systems

Download Center

All assets cryptographically verified
TAR.GZUniversal

LocalMind Python SDK (localmind-sdk)

0.0 MiB

Universal Python 3.10+

SHA-256 CHECKSUM
ebf777c1492937c6ebcc896a9619aeb11d556c2961915bf9f67cc2d70c1d9041
Release Asset Download TAR.GZ

Privacy & Telemetry Pipeline

Loopback-Only Intelligence Flow

100% offline agent orchestration with zero external network leakage.

01 · ENTRYPOINT

Native Rust Launcher

@jarvis sub-millisecond invocation directly to runtime

02 · GATEWAY

127.0.0.1:8420

Strict loopback binding; remote URLs strictly rejected

03 · OBSERVER

Host Telemetry

Deterministic process & workspace indexing without LLM hallucination

04 · STREAMING

Python SDK Events

Typed event stream (token, agent, done) to client terminal

Core Capabilities

Engineered for Reliability

01

Zero Cloud Leakage

Hardcoded loopback-only communication (127.0.0.1:8420); cloud URLs and third-party API calls are strictly rejected.

02

Native Rust Launcher

Builds a zero-overhead binary localmind and symlinks @jarvis into ~/.local/bin without shell script latency.

03

Structured Event Streaming

Emits typed events (meta, agent, token, done, error) with transparent attribution to selected models and system observers.

04

Host Telemetry & Workspace Index

Context-aware agents query live workstation metrics and local codebase files on demand without indexing remote files.

Developer Setup

Installation & SDK Quickstart

Verified tag v0.1.0
Workstation One-Line Install
git clone --depth 1 https://github.com/ravindran-dev/localmind.git "$HOME/.local/share/localmind" && bash "$HOME/.local/share/localmind/install.sh"

Requirement Note: Requires Git, uv, and stable Rust. Builds the native launcher and configures ~/.local/bin/@jarvis automatically.

Private Intelligence

A personal AI agent that lives on your machine

LocalMind provides an autonomous assistant that helps you write code, analyze logs, and run diagnostics without cloud lock-in.

01

Deterministic System Observer

Queries live hardware sensors, active processes, and memory metrics with zero model hallucination.

02

Workspace File Graph

Indexes Git repositories locally to provide instant semantic context to local coding agents.

03

Unified Python SDK

Provides streaming async/sync clients with full event typing, error handling, and timeout safety.

Under the Hood

Technical Specifications

Architecture Blueprint

LocalMind binds exclusively to 127.0.0.1:8420. The native Rust launcher @jarvis forwards commands directly into the locked Python runtime. Local LLMs run via Ollama on loopback:11434, while deterministic system metrics are gathered directly via /proc and sysinfo without LLM hallucination.

Technology Stack
Rust Native LauncherFastAPI / UvicornOllamaPyTorchuvSQLite

Supported Platforms

Linux (x86_64, aarch64)
macOS (Apple Silicon)
Windows WSL2

Implementation Languages

Python 3.12+
Rust
TypeScript

System & Hardware Requirements

Operating System: Linux (Arch, Ubuntu, Fedora), macOS, or Windows WSL2
Python: 3.10 or newer (uv package manager recommended)
Rust: 1.75+ for native launcher compilation
Local LLM Runtime: Ollama or local PyTorch/transformers instance

Developer Documentation

Manuals, Specs & Source

Frequently Asked

Questions & Details

Does LocalMind send any data to external servers?+

No. The SDK and runtime strictly reject any remote URL. All inference runs locally via Ollama or PyTorch on 127.0.0.1.

What is @jarvis?+

@jarvis is the high-performance native terminal CLI launcher written in Rust that provides sub-millisecond invocation for the LocalMind runtime.

Can I use LocalMind as a library in Python?+

Yes! Use the included localmind_sdk library to orchestrate workflows, stream chat tokens, and query system context programmatically.

Official Software Releases

More Products & Developer Toolchains

Explore All Products
Linux Systems

Keynix Keyboard Platform

Active development

System-level evdev remapping, shortcut conflict detection, and typing practice.

Hardware-level Linux input interception using evdev EVIOCGRAB exclusive grabbing and uinput virtual device synthesis, with visual layout editing and shortcut conflict auditing.

tag v0.1.0LinuxRustTypeScript
Desktop & Mobile

Hertix Music Player

Active development

Your music library, kept close and under your control.

A local-first media library and player spanning a native Android app and a Windows desktop app.

tag v1.0.23Android 8+KotlinTypeScript
Developer Tools

Quantum IDE

Active development

A focused, offline-first coding workspace for Windows.

A native desktop IDE built around Monaco, a local Express service and an Electron shell.

WindowsJavaScriptPython