OIOXO in your agents

OIOXO isn't only an IDE. The VS Code extension brings private, local coding intelligence into the editor you already use — a model that runs entirely on your machine, grounded in your codebase — plus the context engine that sits inside the AI tools you already pay for (GitHub Copilot, Claude Code, Cursor, Windsurf, Gemini CLI, Codex) and cuts their token use dramatically.

Private, local coding

• The model runs on your machine. The OIOXO tiers (0.4–4.7 GB, matched to your machine's real memory and GPU) download once, encrypted, and run locally — your code, your questions and the answers never leave the device.
• Grounded in your project, verifiably. Every answer is built on the context engine's capsule — replies show which files they're grounded in, and the tokens that grounding saved you.
• Made for working code. Streamed answers, code blocks with Copy / Insert at cursor / New file, right-click Explain / Improve /Add selection, and Ctrl+Alt+O to open the assistant panel.
• Private image understanding. Attach a screenshot — an error dialog, a UI mockup, a diagram — and OIOXO reads it on your device to ground the answer. Your screenshots never leave the machine; even with a frontier key, only the extracted text is ever sent.
• A frontier model when you want one. OpenAI, Claude, Gemini, Groq, Mistral, xAI, DeepSeek, OpenRouter, a local Ollama, or any OpenAI-compatible endpoint — your key, held in VS Code secret storage, calls going straight to the provider. OIOXO saves your tokens in both directions on every one of them.

The token saver

Agents normally read whole files to answer a question; OIOXO hands them the minimal relevant slice instead.

Install — pick one

• VS Code: install OIOXO from the marketplace. That's the whole setup — when you open a folder, OIOXO wires Copilot, Claude Code and Cursor automatically and shows your live savings in the status bar.
• Terminal (any editor): npm i -g oioxo-mcp, then oioxo-mcp login and oioxo-mcp init in your project. init detects your agents and writes their configs — merged, never overwritten.
• Other VS Code-family editors (VSCodium, code-server, and editors that side-load extensions): download the .vsix and install it via Extensions → ⋯ → Install from VSIX.

What your agent gains

• get_context — the exact code in play plus the API surface of its real dependencies, found by lexical search and the import graph. Called automatically before the agent reads files.
• get_impact — the blast radius of a file: everything that imports it, before a refactor breaks it.
• get_skeleton — a file's full API surface at roughly a tenth of its token cost.
• remember / recall — durable project memory in .oioxo/, shared with the OIOXO IDE — your agent stops re-learning the project every session.

Why it makes answers better, not just cheaper

A model buried in 75,000 tokens of barely-related files misses things. The capsule keeps the signal and drops the noise: anchors arrive as full code, dependencies as signatures only. Less to read, less to get lost in, sharper edits.

Private by architecture

The index is built and queried entirely on your machine — your code is never uploaded. The only thing OIOXO's servers see is the count of tokens you saved, for your meter. Same rule as everything else here: privacy by design.

Plans

Free accounts include a generous monthly saved-tokens allowance — enough to feel the difference on real work. OIOXO Pro ($3.99/mo) makes it unlimited across every agent, every project, every machine. The status bar (and oioxo-mcp status) always shows where you stand.