ApexClaw runs the governance it sells, and publishes what is not working alongside what is.
Omega is an autonomous system operated under ApexClaw's own gate architecture. It has refused unauthorized sends and emitted receipts. It has also earned nothing, completed no full approved-action cycle, and has services that are not running. All of that is below.
Get an Agent Trust Gap BriefWhat Omega is, and is not
Omega is an autonomous system that runs under ApexClaw's own controls. It senses, decides, and attempts actions — and every consequential attempt passes through the same gate architecture ApexClaw sells.
Omega is not a product. It is not for sale, not on a pricing page, and not positioned as something you can buy. It exists here for one reason: a trust vendor that cannot show its own governance working has nothing to show at all.
The record, stated plainly
Refusals on unauthorized sends
The gate has refused attempted external sends that lacked satisfied human approval, and emitted a receipt for each refusal.
Zero real external sends
Across its entire history. Verified at the network layer, not inferred from application logs.
Receipts on gated actions
Decisions and refusals are recorded with the policy version and the check that produced the outcome.
FIRST_PARTY These are our own system's numbers and are labelled as such. They are not market evidence, not a benchmark, and not a claim about anyone else's estate. Last verified 2026-08-04; expires 2026-11-04 — see the claims ledger. A past success receipt is not proof of present runtime health.
What is published with the failures visible
A polished case study proves nothing. What is worth publishing is the part most vendors remove.
- Omega has collected no revenue. It is a governance demonstration, not a commercial system.
- It has completed no full approved-action cycle end to end. The refusals are real; a governed success path with human approval has not yet been run to completion.
- Its brain service is currently masked — deliberately stopped rather than crashed.
- Its signatures are Ed25519, which is not post-quantum. See post-quantum agent evidence.
- Several dependent services are not running. They are listed as NOT_OPERATIONAL internally rather than described as almost ready.
Publishing that list is the point. The governance claim is "actions are refused unless authorized, and there is a receipt either way" — and that claim survives every item above, because none of them is a claim we made and failed to meet.
The vocabulary this system is held to
Internally, status may only be described using: PROVEN · CANDIDATE · PLANNED · HISTORICAL · CONFLICTED · UNKNOWN · UNTESTABLE · PASS · FAIL · NOT_OPERATIONAL.
Banned: LIVE · OPERATIONAL · PRODUCTION READY · "almost done" · "should work" · any percentage-complete. Those words let a system describe itself as further along than its evidence supports, which is precisely the failure this whole layer exists to prevent.
Common questions
What is Omega?
A governed autonomous system operated by ApexClaw under ApexClaw's own controls. It exists here as evidence, not as a product.
Why publish it as proof?
Because running your own governance against a real autonomous system is the one credential that cannot be claimed without doing it.
What do the receipts show?
Actions attempted, policy decisions, approvals, and refusals — including runs where the correct outcome was that nothing was sent.
Are the numbers audited?
No. They are first-party with stated methodology. Where a claim has not been independently verified, it says so.
Can we see a live run?
Sample artifacts are published under /.well-known/. Anything beyond that is arranged directly.