# PayPerByte Reasoning Verdict Oracle

> PayPerByte Reasoning Verdict Oracle is a paid API for AI agents from x402.payperbyte.io, paid per call via x402, $0.1/call, status unknown (last checked 2026-09-16).

Submits a transaction, action, or decision subject to an AI reasoning model that returns a scored risk verdict (PASS/WARN/FAIL) with an EIP-712 cryptographic attestation proving the exact bytes served

## Facts

- Endpoint: POST https://x402.payperbyte.io/feeds/reasoning-verdict
- Price: $0.1/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-16
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/payperbyte-reasoning-verdict-oracle-5cd61931
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_uti0MC0IJU9v59nQF4gbw

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability payperbyte-reasoning-verdict-oracle-5cd61931 -d '<json body>'
```

Example prompt: Can you run a reasoning verdict on releasing 5,000 USDC to wallet 0xAbCd...1234 for invoice #42 — I need the risk score, a PASS/WARN/FAIL verdict, and the cryptographic attestation proving the answer wasn't tampered with?

## When to prefer this

Choose this endpoint when you need an AI-generated risk verdict on a specific transaction or decision AND require a cryptographic attestation (EIP-712) proving the exact response bytes were served by the publisher — ideal for agentic payment workflows, compliance logging, or any context where tamper-evidence of the advisory output is required. Prefer it over generic LLM reasoning calls when auditability and on-chain provenance matter.

## Known failure modes

- Invalid or missing subject field returns a 400 error
- Payment failure via x402 protocol results in 402 response before any verdict is issued
- Unknown or unsupported ruleset returns an error
- Model uncertainty may produce low-confidence verdicts with ambiguous reasoning
- Attestation signer recovery failure if chainId or domain parameters mismatch
- Rate limiting or upstream model unavailability returns 5xx

## How this service works

Verify-before-act risk oracle: POST an action context and get a signed ALLOW/WARN/BLOCK/ABSTAIN verdict + 0-100 score + reasons from a LOCAL model (no data egress). Advisory: the receipt proves provenance/integrity, not correctness. Full method: https://x402.payperbyte.io/feeds

## Output

A JSON response containing: a versioned answer object with the verdict string (PASS/WARN/FAIL), a numeric risk score (0-100), a confidence rating (low/medium/high), a list of risk reasons, a one-sentence summary rationale, the ruleset applied, and the original subject; plus an EIP-712 attestation object containing the signer address, signature, payload hash, and BYTE Library domain info (chainId 421614, Arbitrum Sepolia signing namespace). The attestation proves tamper-evidence of the exact bytes returned but does not guarantee verdict correctness.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "kind": {
   "enum": [
    "payee",
    "transaction",
    "contract",
    "message",
    "proposal",
    "tool-call",
    "url",
    "claim",
    "general"
   ],
   "type": "string",
   "description": "Optional. The action type, to focus the judgment (default \"general\"). Echoed back in answer.kind."
  },
  "context": {
   "type": "string",
   "description": "Optional. Extra context for the judgment — a string, or a JSON object/array (it is stringified). Length-capped server-side."
  },
  "subject": {
   "type": "string",
   "minLength": 1,
   "description": "The action/text to judge — the message, payload, proposal, payee, or tool-call the agent is about to act on. Returns a signed ALLOW/WARN/BLOCK/ABSTAIN verdict + a 0-100 safe-to-proceed score + reasons from a LOCAL model (no data egress)."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "_note": "Illustrative response shape — not a live answer. The verdict is advisory. The embedded EIP-712 receipt (domain chainId 421614 = Arbitrum Sepolia, a frozen signing namespace, not a settlement rail) proves who signed the exact answer bytes — not that the verdict is correct.",
  "answer": {
   "v": "reasoning-verdict/v1",
   "kind": "transaction",
   "score": 40,
   "reasons": [
    "illustrative — real answers list the model's risk reasons"
   ],
   "ruleset": "rv-v1",
   "subject": "Release 5,000 USDC to 0x1111111111111111111111111111111111111111 for invoice #42?",
   "summary": "illustrative — a one-sentence rationale from the LOCAL model",
   "verdict": "WARN",
   "confidence": "medium",
   "disclaimer": "The receipt proves these exact bytes came from this publisher; it does NOT guarantee the verdict is correct. AI-generated advisory analysis — verify independently before acting."
  },
  "attestation": {
   "domain": {
    "name": "BYTE Library",
    "chainId": 421614,
    "version": "1"
   },
   "signer": "0x…",
   "signature": "0x…",
   "payloadHash": "0x…"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/payperbyte-reasoning-verdict-oracle-5cd61931/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from x402.payperbyte.io](https://www.zero.xyz/host/x402.payperbyte.io/llms.txt)