Get Started
Cube v1.0.0•Engineered in the Terminal

A coding agent that lives where you already work

Differential-rendering TUI, multi-mode authentication (Claude Pro, ChatGPT Plus, 15+ API keys, or offline Ollama), workspace memory, and atomic code mutations.

>irm https://cube-agent.pages.dev/install.ps1 | iex
SHA-256 Verified Release•Windows x64•
0developers installed across Windows, MacOS & Linux
STANDARDIZED AGENT HARNESS SPECIFICATION

The 6 Pillars of the Agent Harness

The operational runtime scaffolding that elevates raw intelligence into an autonomous software engineer.

01
LOOP
Loop
COMPONENTACTIVE
02
TOOL
Tool
COMPONENTSTANDBY
03
CONTEXT
Context
COMPONENTSTANDBY
04
MEMORY
Memory
COMPONENTSTANDBY
05
BACKGROUND
Background
COMPONENTSTANDBY
06
GUARDRAIL
Guardrail
COMPONENTSTANDBY
Pillar 01 · LOOP
Iterative Decision Cycle
01 / 06

Loop

The central execution runtime. An agent harness runs a continuous loop that observes the environment, plans an action, executes it, and evaluates the result — repeating until the task is done.

Feedback-Driven Self-Correction:Reads back errors, test failures, and diagnostics from the environment and adjusts course automatically, instead of guessing blind.
Bounded Execution:Enforces turn limits and cycle detection so the agent can't loop forever or burn through resources on a stuck task.
Live Progress Streaming:Surfaces reasoning and actions as they happen, so you can follow along in real time instead of waiting on a black box.
Observe-Plan-Act-EvaluateSelf-CorrectionCycle GuardrailsReal-Time Streaming
CORE CAPABILITIES

Terminal intelligence engineered for developer flow

Engineered for speed, privacy, and full developer agency. Explore how Cube orchestrates models, workspace context, and safe execution directly inside your shell.

GW
SUB
SKILL
Q&A
HITL
MCP
FACE 1/6 · FRONT
GATEWAY

Multi-gateway model routing & OAuth

~/your-project | gpt-6-astra
FRONT
01 · FRONT FACE: GATEWAYFacet 1 of 6

Multi-Gateway Model Matrix

Your subscriptions. Your keys. Zero lock-in.

Connect directly to Claude Pro, ChatGPT Plus, or Grok memberships via browser PKCE OAuth 2.0 with zero token markup. Plug in any of 15+ developer API keys, or run 100% offline with local Ollama models.

  • ›PKCE OAuth for Claude Pro / ChatGPT Plus / Grok
  • ›15+ Developer API providers with unified fallback routing
  • ›100% Air-gapped offline inference via Ollama & LM Studio
COMMAND INTERFACE

Command-line precision built for flow

Keep your fingers on home row. Instant slash commands give you total control over models, memory compression, and thread branches without interrupting your work.

cube ~ /model
Model & GatewaySyntax: /model [model_name]

Switch the active reasoning model instantly without restarting your agent session.

❯/model claude-3-7-sonnet
Terminal Output:
Switched active model:
  From : claude-3-5-sonnet-20241022
  To   : claude-3-7-sonnet-20250219 (Thinking enabled)
Context preserved across 8 turns.
KNOWLEDGE BASE

Frequently Asked Questions

Architecture, data privacy, subscriptions, and getting started.