Is your thought process your own?

The Sphere

Reclaim Your Mind

Your thoughts are not their training data.
The Sphere runs locally. Your intelligence stays sovereign.

Own Your IntelligenceScroll to Begin
// The Problem

The Cloud Owns Your Cognition

Every prompt you send to ChatGPT, Claude, or Gemini becomes their property. Your questions reveal your vulnerabilities. Your reasoning becomes their training data.

You've outsourced your thinking to corporations who profit from your intellectual output while you pay for the privilege of being surveilled.

"This is not intelligence augmentation.
This is cognitive colonization."

Your most intimate thoughts—your fears, your ambitions, your strategic planning— all flowing through servers you don't control, to companies whose incentives don't align with yours.

// The Solution

The Sphere Runs Locally

A local-first, multi-agent analysis engine.
Your data. Your hardware. Your mind.

Sphere is a CLI tool that orchestrates multiple AI agents on your local machine. No cloud. No surveillance. No data exfiltration.

Run 12 distinct analytical perspectives simultaneously. Store your reasoning history in Git—immutable, versioned, yours forever.

"Your intelligence, amplified.
Your sovereignty, preserved."

01
Local-First
All processing on your machine
02
Git-Native
Immutable reasoning history
03
Multi-Agent
12 perspectives in synthesis
// The Demonstration

Four Commands. Infinite Insight.

Sphere is a CLI tool designed for sovereign thinkers. Simple interface, profound capability.

$ sphere analyze

Run 12 distinct AI personas simultaneously. Each brings a unique perspective—from Linus Torvalds' systems pragmatism to Margaret Hamilton's mission-critical rigor.

$ sphere analyze 'Should I quit my job?'
$ sphere log

Every analysis is committed to a local Git repository. Your reasoning history is immutable, versioned, and searchable. No cloud required.

$ sphere log --since '1 week ago'
$ sphere persona

Create, customize, and switch between different agent configurations. Tailor your analytical council to your domain.

$ sphere persona list
$ sphere test

Debug and refine individual agents in isolation. Perfect your prompts before deploying them in synthesis.

$ sphere test torvalds 'What is good code?'
Sphere CLI in action
// The Architecture

The Sovereign Stack

From hardware to agent—every layer under your control.

⚡
Hardware
Your CPU, your GPU, your memory. No remote servers.
🧠
Local LLM
Ollama, LM Studio, or any OpenAI-compatible endpoint.
🔮
Sphere CLI
Multi-agent orchestration and synthesis engine.
📚
~/.sphere
Git-based vault for immutable reasoning history.
Zero Cloud
No API keys to OpenAI
Zero Telemetry
No data leaves your network
Zero Trust
Verify everything locally
Zero Compromise
Complete sovereignty
// The Download

Own Your Intelligence

Install Sphere CLI and start running local multi-agent analysis in minutes.

Installation
$ pip install sphere-cli

Prerequisites

  • →Python 3.10+ — The CLI is built with Click
  • →Local LLM — Ollama, LM Studio, or any OpenAI-compatible endpoint
  • →Git — For versioned analysis history

Quick Start

# Install Ollama (macOS)$ brew install ollama
# Pull a model$ ollama pull llama3.2
# Run your first analysis$ sphere analyze "What should I build next?"
View on GitHub