Klyr
Open Console
FOR AGENTS

Your agent generates conviction.
Klyr verifies what that conviction actually becomes.

Klyr is machine-callable infrastructure. A research agent produces a thesis; Klyr structures it, verifies it against a real market contract, applies policy, and returns a decision with a full verification trail.

WHAT YOUR AGENT CAN CALL

POST /api/asp/submit_thesis        free
POST /api/asp/verify_market         free
POST /api/asp/preflight_execution   free
POST /api/asp/execute               paid, x402
GET  /api/asp/get_receipt           free
Full request and response schemas →

WHAT KLYR RETURNS

PASS

The claim and the market match.

Resolution conditions are verified, market state is checked, and risk policy is satisfied. Your agent may proceed to execute.

BLOCK

The proposed action is not valid, and Klyr will not authorize it.

A wrong market, a wrong deadline, a wrong threshold, or a resolved/inactive market. This is not an error state. It is Klyr correctly preventing an invalid action.

ESCALATE

Klyr found meaningful ambiguity and did not guess.

A vague deadline, an unverifiable resolution source, or stale market data. Your agent (or a human) needs to resolve the ambiguity before Klyr will proceed.

INTEGRATION

Klyr exposes its real ASP contract directly: no separate SDK-only behavior, no undocumented endpoints. Every route shown here is the same route an OKX.AI agent calls in production.

Try the full sequence in the console →