Skip to content
AIContinue
Developer Preview · SDK 0.3.0.dev2SDK available for development and pre-certification. Official certification requires separate review.

AICONTINUE / DEVELOPERS

Build adapters.
Test locally.

A bounded capability contract for Software, Device and Hybrid adapters. Start with synthetic examples, then prepare evidence for review.

What changed in 0.3.0.dev2

One CapabilityInterface for Software, Device and Hybrid, with optional motion enabled only by explicit declarations and permissions. Fixes isolate rejected requests from authorized events, enforce consistent state/error results, and include the approved license in the wheel.

59 local checks passed, 0 failed, 5 blocked; 4 examples passed. Physical safety, measured hardware safe-state, real transport, authenticated Core integration and certification remain pending. This preview does not control real robots. Install in a dedicated Python environment.

HARDWARE DEVELOPERS

Bring your product into the AIContinue ecosystem.

Start with the public capability contract and offline examples. Integration into the AIContinue Certified program and use of its seal require explicit AIContinue approval for the product/model, version and capabilities.

  1. 1

    Explore the adapter and test locally

    Read the quickstart, capability limits and simulator workflow. Download SDK 0.3.0.dev2 and run the examples and pre-certification in a local Python 3.11+ environment.

  2. 2

    Tell us about your hardware

    Introduce your product/model, intended capabilities and dependencies through the existing support contact. Keep confidential source code, credentials and personal data out of the first message.

    Use the existing support contact for an initial conversation. There is no live certification submission form here.

  3. 3

    Prepare for review

    Use the manufacturer checklist and local reports to prepare evidence. Real hardware/security validation, production integration and seal use remain separate approval gates.

What AIContinue Certified means

Documentation

English and Simplified Chinese technical documentation. AI technical review is complete; no external human review has been performed.

Local checks are a starting point.

The simulator uses synthetic data. No physical device, production service or provider credentials are required.

LOCAL EVIDENCE

What you can test

  • Declared capabilities and bounded request/response shapes
  • Synthetic examples and simulator behavior
  • Local conformance and pre-certification report generation

FURTHER REVIEW REQUIRED

What still needs approval

  • Physical hardware, firmware and vendor SDK behavior
  • Security review, permissions and production integration
  • Official certification, credentials and use of a certification mark

A local result, including ReadyForCertification, means only that the submitted evidence may be prepared for offline review. It does not grant AIContinue Certified status, production access or permission to use a certification mark.

Prepare a certification review request

This is a documented review process. There is no active submission form or certification service on this page.

  1. 1

    Build the evidence packet

    Include adapter identity, version, declared capabilities, hardware and firmware scope, local reports and known limitations. Use sanitized samples.

  2. 2

    Confirm the official review channel

    Wait for an AIContinue-approved intake route and sharing terms before sending evidence. Do not send credentials, personal data or confidential source code through this site.

  3. 3

    Complete independent review

    Hardware and security validation, a scoped decision and authorized issuance remain separate gates. The offline toolkit cannot approve them.

Read packet requirements