> ## Documentation Index
> Fetch the complete documentation index at: https://documentation.orbitdev.org/llms.txt
> Use this file to discover all available pages before exploring further.

# Identity and purpose

> Orbit AI's identity, mission, long-term vision, and core principles for building a modular, practical, and accountable AI platform.

## 2.1 What Orbit is

Orbit AI is the product and research vision. **Orbit Labs** can serve as the name for the research and development identity, while a formal legal entity and its exact registered name should be stated only after it has been established and verified. Brand names, product names, and legal company names are not automatically interchangeable.

The Orbit ecosystem is intended to bring together several layers:

1. **Experience:** the interfaces people use to interact with AI.
2. **Intelligence:** models that interpret instructions, generate content, and reason over supplied context.
3. **Action:** tools and integrations that let an approved model perform defined tasks.
4. **Infrastructure:** the services that manage requests, identity, data, model execution, and observability.
5. **Governance:** policies and technical controls that keep the system secure, reviewable, and under human direction.

The layers should remain modular. The user experience should not depend on a single model provider, and a model should not receive unrestricted access to every connected account simply because a user asked it to help.

## 2.2 Mission

**To make advanced AI practical, accessible, and dependable by connecting capable models to the tools and workflows people use.**

The mission has three parts:

* **Practical:** solve real tasks rather than optimize for impressive demonstrations alone.
* **Accessible:** make complex capabilities understandable to people with different levels of technical experience.
* **Dependable:** build around transparent limitations, robust security, and measurable performance.

## 2.3 Long-term vision

Orbit's long-term vision is an integrated AI platform that can support people across digital work and, eventually, carefully bounded physical tasks. That vision may include more capable models, agentic software, multimodal interaction, and robotics. Each step must be justified by evidence, engineering readiness, and risk assessment.

Artificial general intelligence is a research ambition, not a label that should be applied casually to a product. No marketing phrase can establish that a system is generally intelligent. Claims about advanced reasoning, autonomy, or generality must be backed by evaluations that are relevant, repeatable, and disclosed with their limitations.

## 2.4 Principles

**Utility over spectacle.** A feature should make a task better, faster, safer, or more accessible.

**Evidence over hype.** Public claims must match what has actually been built and tested.

**Human direction.** People should understand what an agent is allowed to do and how to stop it.

**Security by design.** Permissions, isolation, logging, and recovery are architectural concerns, not finishing touches.

**Modularity.** Models, interfaces, and integrations should be replaceable where practical.

**Respect for users.** Data collection should be limited, transparent, and justified.

**Measured ambition.** Long-term research is encouraged, but releases should be based on readiness rather than deadlines alone.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.