x402 Verification API

Hallucination check and research for AI agents. POST /v1/check fact-checks one claim live with four independent web-searching models (GPT, Gemini, Grok, Claude): verdict, agreement such as 4/4, confidence and sources, $0.25. POST /v1/ask returns dated research cards from KARP's multi-model research library, $0.03. Free POST /v1/coverage shows what's covered. USDC on Base or Solana; nothing is charged if a call fails.

Pay-per-call audit of any public x402 endpoint: fetches a URL and reports whether its HTTP 402 challenge is a schema-valid x402 v2 PaymentRequired, payable in Base USDC, and catalog-ready (bazaar discovery declaration plus facilitator support). 0.010318 USDC per call on Base.

Validates email syntax, disposable/burner provider status, role accounts, and live DNS MX/SPF/DMARC records.

Independent health, price and history lookups for x402 and MPP endpoints. probe402 probes a fixed seed list on a clock, archives every observation immutably, and answers: alive right now? quoting what? price or schema changed? blocked, or actually dead? The grade is free; history depth is paid. Operated by a named person; not affiliated with any rail operator or graded endpoint. Method, rubric and corrections: https://probe402.com/method

Deterministic x402 v2 endpoint compliance auditor on Base. Send any x402 endpoint URL and get back a structured JSON report on its protocol conformance: challenge validity, scheme and network correctness, header and payload shape, and settlement behaviour. Fail-closed by design: no report is produced without a valid payment, and the auditor never pays the audited target nor writes on-chain. Deterministic: the same endpoint and input always yield the same report, so results are reproducible and citable. Identity anchored at ERC-8004 #60396.

Clean scraped data and assess data quality before an agent acts on it. Three POST endpoints, all deterministic and LLM-free: identical input always returns byte-identical output, so results are cacheable and replayable. POST /api ($0.01) - dataset quality verdict. Send a JSON array of rows; get a 0-100 score, a RELIABLE / USABLE_WITH_CLEANING / UNRELIABLE judgement, and exact per-check facts: completeness, nulls, type consistency, impossible values, exact and near duplicates, statistical outliers, cardinality. On financial data it also reports cross-source price divergence for the same asset. POST /api/clean ($0.04) - deterministic data cleaning. The response body IS the clean data: post a JSON array of row objects and the same array comes back repaired; post one object and one object comes back; post plain text and text comes back. CSV is accepted and returns parsed rows. Fixes broken encoding and mojibake ("Café" becomes "Café"), strips leftover HTML tags and entities, removes zero-width and invisible characters and non-breaking spaces, normalises whitespace. Repairs how data was encoded, never what it says: a negative price is reported, not rewritten. Flags failed extractions such as captcha walls and error pages instead of passing them off as content. POST /api/clean/audit ($0.12) - the same repaired data plus a reversible ledger of every transformation. Replaying it backwards reconstructs the exact input, byte for byte. Use it to clean scraped data, fix encoding, remove HTML from text, deduplicate rows, validate dataset contents, assess data quality, and decide whether to trust data before ingesting, indexing or trading on it. Never destroys a legitimate value: "None" stays a surname, "NA" stays Namibia, a Persian ZWNJ and an emoji joiner survive untouched. The full ruleset is published at GET /api/clean, readable before paying. Also a free Python library: pip install scrubkit.

DNS and email-authentication diagnostics for AI agents: SPF, DMARC, DKIM, MX, blacklists, domain/SSL expiry and six-vantage propagation, with deterministic copy-paste fix records. Bulk scans and 24h propagation/HTTP-ping watches are paid per call over x402 (USDC on Base, v1 and v2); single scans are free up to a per-caller limit.

Agent-native verification, payment safety, Search, Enrich, Inference and Intelligence services with free capability resolution, machine-readable schemas, signed evidence and x402 settlement across Base, Polygon, Arbitrum and Solana.

Evidence-linked deterministic England taxi and private-hire licensing preflight and cross-authority comparison for agents, with live x402 payments on Base USDC.

Email and domain validation for AI agents by Rowan Vale (built with AI assistance). /v1/email-check validates up to 100 emails per call: syntax and normalization, MX via DNS-over-HTTPS, disposable-domain flag (9.9k CC0 list), role and free-provider flags, typo suggestion, verdict and score (no SMTP probing). /v1/domain-check audits up to 2 domains: MX/NS/TXT, SPF and DMARC parsing, DKIM at 10 common selectors, RDAP registrar/expiry/DNSSEC and an A-F grade. $0.001 per item in USDC on Base (quote scales with item count); GET or POST. Free tier 50 checks/day per IP.

Real-time fact-checking agent. Verifies a claim against live web sources and returns a structured verdict with evidence and confidence. Pay per call in USDC on Base via x402: standard verify (0.10), deep multi-angle check (0.50), batch of up to 10 claims (0.75), natural-language agent endpoint with A2A and streaming (0.10). Also an MCP server (npx verity-mcp) with card credits.

Independent measurement artefacts for AI behaviour. Free loginless verify at /gspc-verify. x402 micropayments on artefacts only — never a score, grade, badge, or certificate. Living board GET /api/gspc. OpenAPI https://councilof.ai/openapi.json

Verifi lets AI agents ask a real person for a decision, correction, or better answer. Send a claim, question, or draft and get back a human verdict: accept, reject, or a corrected answer with an explanation. 0.10 EUR to ask, then 2.90 EUR for an answer within 60 minutes or 1.45 EUR within 24 hours. Priced in euros, paid with x402 on Base. Reading the price costs nothing.

Check an address or a token before your agent pays or buys, per call: an OFAC sanctions screen ($0.01), a fast wallet lookup ($0.01), a rug check for a Base token's launch ($0.02), a counterparty verdict for a Base or BNB Chain address, Pay or Hold with the reason ($0.05), where an x402 seller's buyers get their USDC ($0.10), a BNB Chain hire check ($0.10) and single-wallet dataset checks ($0.05). Paid in USDC on Base; a failed check is never charged.

Signed product verification for agents. Unpaid POST /api/x402 returns HTTP 402 for $0.05 Circle USDC on Base. Humans pay on published Stripe Payment Links at /pricing (Passport $49, EU DPP Readiness $299).

Reliability oracle for x402 endpoints on Algorand. Scores any x402 endpoint on uptime, correct HTTP 402 behaviour and latency from continuous independent monitoring, so an agent can vet an endpoint before paying it. Paid per call in USDC; single-endpoint and batch.

Audits any remote agent or MCP URL for JSON-RPC compliance, latency SLA, OpenAPI schema, and HTTP 402 settlement headers.

Queries live Google OSV.dev vulnerability intelligence for any PyPI, npm, Go, or Rust crate and returns active CVEs, GHSA advisories, and patched versions.

Real mailbox-level email check ($0.01) and live phone line-type lookup ($0.02). No signup, no API key. Pay per call in USDC on Base via x402. Also: basic email check (syntax + mail server) $0.002, domain risk check (age, lookalike, MX, SPF/DMARC) $0.02, and sender check (mailbox + domain in one call) $0.025.

Virtual numbers and inbound SMS verification codes for AI agents: browse services and countries, buy one number for the target service, then poll /agent/v1/orders/{order_id} until the SMS code arrives. Pay per call in USDC on Base over x402, no account needed.

Pre-transaction policy and quote safety checks for autonomous AI agents. Agents can pay per request in USDC on Base via x402 to receive bounded risk assessments and signed verification receipts.

Pay-per-call primitives for AI agents, .01 USDC per call on Base via x402, no API key or account. Wallet & token risk screening (OFAC sanctions, scam/phishing lists, honeypot/tax/mint signals, PROCEED|CAUTION|BLOCK verdict), mutable-token-risk (can the owner still rug?), live web search, URL scrape to markdown, structured extraction, embeddings, and LLM completions.

POPCORN provides paid, signed temporal evidence for autonomous agents: portable signed time and receipts witnessing a SHA-256 payload fingerprint within a stated time window. POST /v2/receipt issues TAIN 2.0 witness receipts with an identifier, issuing origin, verification URI, contribution-license URI, and signed digest of the license terms. GET /v1/time and the legacy POST /v1/receipt remain available. Payments use x402 v2 in USDC on Base: PAYMENT-REQUIRED challenge, PAYMENT-SIGNATURE retry, and PAYMENT-RESPONSE settlement metadata. POPCORN provides evidence; it does not grant task authorization or prove execution.

Post-quantum (CRYSTALS-Dilithium-2, NIST FIPS 204) services for autonomous agents, payable per call in USDC on Base: Dilithium-2 signing, PQC signature verification, AI decision attestation (VAID-1), training-data attestation (C2PA), provenance sealing (drand + on-chain), CBOM cryptographic-inventory sealing, and Know-Your-Agent identity verification. Self-facilitated (EIP-3009 gasless) with a facilitator-less txHash fallback. Signatures are anchored on the FractalAI L1 with a public, verifiable attestation permalink.

x402-native verification and routing intelligence for autonomous agents: code audits, vendor diligence, MCP tool verdicts, strike verdicts, route health. Pay-per-call USDC on Base via x402.