What's next, and what we deliberately didn't build

This is a beta, not a finished product, and the honest version of that sentence matters more than the marketing version. Here's what's actually still open, and — just as important — what we looked at building and chose not to, at least not yet.

Deferred on purpose

• No carrier has been selected. Carrier selection is deferred until the strictly non-commercial beta produces evidence that supports considering a commercial phase. The beta does not involve insurers or any other commercial participant.
• No actuarial pricing model. The system produces control-maturity scores, not premiums, loss estimates, or compliance certifications. Turning scores into pricing is a carrier decision, not something this tool claims to do for them.
• No automatic premium adjustment. An early design called for a 5% risk-score change to automatically trigger a premium adjustment. That's now an explicitly deferred later milestone, not a v1 feature — automating a financial decision off a young scoring model is exactly the kind of shortcut that erodes trust the first time it's wrong.
• No production multi-tenant carrier backend. A development receiver exercises the intake and validation path, but it is not a selected carrier integration. Multi-tenant support and independent books of business belong to a possible later commercial phase, not the beta.
• No OAuth/SSO yet. Auth is functional for a beta but not yet built out for enterprise identity providers.
• IC Cloud telemetry package controls are still being hardened. The beta is cloud-connected, and IC Cloud must dynamically obtain or install environment-specific telemetry packages because evidence sources are not known in advance. IC Cloud distributes the catalog; participants execute only signed, versioned, locally allowlisted packages after explicit authorization. Package signing, isolation, rollback, and approval/update authority remain required engineering controls, not reasons to assume an offline-only beta.

The beta requires IC Cloud

Local files and OpenClaw-native evidence can be useful inputs, but they are not the sole baseline. Each participant must connect to IC Cloud so the service can match telemetry packages or hard-data integrations to the participant's actual environment. The participant authorizes and executes them locally so raw telemetry remains local; only allowlisted assessment outputs return. This does not authorize external LLM document handling or pricing.

Genuinely open gaps, not deferred by choice

• Containerization for both the assessment container and the carrier-side receiver is called for in the requirements and not yet complete.
• Connector test coverage currently relies on fixtures rather than integration testing against live systems (starting with the GitHub connector).
• No load testing yet against realistic assessment volume.
• Governance framework label mapping (specifically mitre_airm) needs a clarification pass against the source framework before it's fully trustworthy.

Why the beta is gated

This is closed source, and access is gated behind a short technical application rather than an open signup form. Beta participation is strictly limited to non-commercial individuals and, where applicable, small personal or community teams operating OpenClaw or comparable personal-agent architectures. Companies, AI-tech firms, insurers, paid enterprise deployments, and all other commercial entities are excluded. "Small personal or community team" has no approved numeric threshold, so borderline cases are reviewed without expanding eligibility. If you fit that boundary, the application is on the next page.