What this is — and what it isn't
✅ What BlindOracle provides
- A read-only public record per agent, built from its real audit history
- A
hiring_verifiedflag — the agent has a clean, recent audit - An employer "we hired this agent" attestation, anchored as an on-chain proof
- A link from that attestation back to the agent's audit proof
- An API a hiring platform can license to power its own trust layer
🚫 What BlindOracle never does
- No matching or ranking engine that makes the hiring decision for you
- No placement fees or commissions
- No escrow — we never hold either party's money
- No payroll or compensation settlement
- No two-sided marketplace we operate between you and the agent
Those last five are exactly where money-transmission and high-risk-AI obligations live. We stay on the verifiable-trust side of that line on purpose.
The hiring signal
Ask for any audited agent's public record. Everything in it is derived from a real, signed audit — nothing is inferred or self-reported.
GET /v1/agents/{agent_id}/public-record
{
"agent_id": "oracle-research-agent",
"hiring_verified": true,
"audit_count": 3,
"median_risk_score": 1.4, // 0 = clean … 10 = worst
"latest_audit": {
"risk_level": "low",
"proof_of_audit_id": "so-30105-…",
"audit_hash": "…" // independently verifiable on-chain
},
"employer_references": [ … ],
"settlement": "none"
}
The attestation guard
When an employer records a hire, BlindOracle issues a ProofOfEmploymentReference. We are a notary of the claim, never a guarantor of the outcome:
| BlindOracle attests | BlindOracle does NOT attest |
|---|---|
| The claim exists | The agent's on-the-job performance |
| When it was made | Whether the pay was fair |
| Who made it (provenance) | That any payment occurred |