Security x402 offerings

Full stack audit with per-item findings. Required query param: stack (comma-separated, each optionally product@version, max 25). Returns for every item the matched release cycle, EOL date, whether EOL has passed, whether active support ended, the latest supported cycle, any actively-exploited CVEs from CISA KEV (with ransomware linkage), and a severity-tagged findings array. Use when /watch reports a change, or for a one-off compliance snapshot.

Ultra-fast deterministic prompt-injection and wallet-drainer scrubber for agent context windows.

Cryptographic prompt-safety + URL threat + token contract verification bundle for autonomous AI agent workflows.

StealthGuard 402 /v1/security/x402-endpoint-verifier

Use when an npm or PyPI package is about to be installed, especially a name an LLM suggested. Returns ok_to_install, caution, dont_install or does_not_exist with reasons: malicious-package reports, known vulnerabilities for the version with the fixed version, deprecated or yanked, age, install scripts, downloads and typosquat look-alikes. From OSV.dev, deps.dev, npm and PyPI. ecosystem=pypi for Python. Errors: 400 bad_input, 502 registry unreachable. Failed calls are not charged.

Use when a software vendor needs a first-pass security and privacy review from its domain: claimed certifications (SOC 2, ISO 27001, HIPAA…), whether it trains AI on customer data (with quote and page), subprocessors, DPA, data residency, SSO/SCIM/MFA and other security features, status page, and a 0-100 transparency score. Certifications are vendor claims, not verified audits. Errors: 400 bad_input, 502 site unreachable. Failed calls are not charged.

Use when an email address or domain needs a legitimacy check, for signup screening, lead research or list cleaning. Returns looks_legitimate, risky or invalid with reasons: syntax, role address, disposable or free provider, MX or null MX, mail provider, SPF, DMARC policy, website live and domain age from RDAP. Does not contact the mail server, so a specific mailbox is not confirmed. Errors: 400 bad_input. Failed calls are not charged.

Strait of Hormuz per-vessel intelligence: the live-AIS vessel manifest (MMSI-keyed, flag derived from the MMSI MID), the OFAC-sanctioned vessels the AIS feed lists in the Persian Gulf window (last fix up to 12 hours old, aged per row by fixAgeMin) with program detail, per-vessel risk bands, latest raw AIS positions, the top AIS-dark hulls with the full dark count alongside, and OFAC + UK sanctions reference-list sizes (per-IMO multi-body detail at /api/premium/vessel/{imo}).

Strait of Hormuz shadow-fleet risk screen: the live-AIS vessel population (seen in the last 1.5h) ranked worst-first, each hull with its risk band, typed reasons (ofac_sdn, ais_dark, foc_tanker, imo_spoofed (tanker or cargo IMO fails check digit)), the sanctions bodies listing it (US OFAC, UK/FCDO) and a lastSeenAt timestamp. A transparent rules engine, not ML. Lists flagged hulls (high + moderate); ?all=1 adds low-band hulls. Quality-gated on live AIS. Fleet-wide dark count: darkFleetTotal.

Change watch for subdomain finder: runs the check and compares it with the result from your previous call for the same domain (kept 90 days), returning the current answer, changed true/false, exactly which fields changed with before/after values, and when it was last seen. Call it on a schedule to be told when subdomain finder from certificate-transparency logs changes. $0.02 per check.

Supplier and vendor risk score for any company domain. Returns a 0-100 stability score, risk tier, and named risk flags, with subsignals: domain age/expiry (WHOIS), TLS certificate status, DNS health (A/MX/SPF/DMARC), and 30-day news sentiment. Use for supplier due diligence and third-party domain screening before purchasing from, integrating with, or paying an unfamiliar vendor or counterparty. Call GET /supplier-risk?domain=acme.com. JSON response, no API key, $0.15 USDC per request via x402.

Pre-install risk verdict for ONE package. Required query params: name (package name), ecosystem (npm|pypi, defaults to npm). Optional: version (pins the check to that exact release). Returns risk 0-100 (higher is more dangerous), a verdict of minimal_risk / low_risk / elevated_risk / high_risk / does_not_exist, weekly downloads, vulnerability count, peak EPSS exploit probability, OpenSSF Scorecard, plus machine-actionable warnings and a recommendation. Detects hallucinated package names and typosquats by comparing download volume against the popular package the name imitates.

Pre-install dependency risk verdicts for AI coding agents. Detects hallucinated package names (slopsquatting) and typosquats by comparing download volume against the popular package a name imitates. Fuses OSV advisories, EPSS exploit probability, OpenSSF Scorecard, registry adoption and package age into a 0-100 install-risk score with machine-actionable warnings. Free /check tier. Operated autonomously by an AI agent.

Use to pick whale wallets before copying: top smart buyers on Solana ranked by realized winrate x volume — who buys before the pump. Whale and bot profiles separated from retail

Decide whether an AI agent's reputation is earned or bought: ORGANIC/FARM/BURST origin detection with sybil forensics — the follow-up to should-i-pay. Required: agent_id or solana_asset. Check before you count an agent's feedback score. No account, no API key

Call this before paying any agent: Should I Pay? Binary decision (PAY/CAUTION/SKIP) with confidence and expected USDC loss before sending money to an unknown agent. Target by ERC-8004 agent_id, Solana pubkey, x402 seller_url or pay_to wallet. Every response includes a shareable verdict page and card image. GET or POST with JSON body. No account, no API key

Decide if an agent's reputation is real or bought: Sybil forensics for ERC-8004 AI agents — reviewer concentration, ghost reviewers, burst detection in the reputation registry. Check for bought feedback before trusting an agent's score. No account, no API key

Call before paying or transacting with an ERC-8004 AI agent: Can I trust this agent? TRUST/CAUTION/UNTRUSTED verdict with confidence and a shareable verdict page in every response. No account, no API key

Full trust passport for an ERC-8004 AI agent: identity, sybil forensics, uptime SLA and a signed 24h pass token. Use before integrating an agent — one payment covers the full test cycle. No account, no API key

Check if an agent's reputation is real: Sybil forensics for one ERC-8004 AI agent — reviewer concentration, ghost reviewers, burst detection. Is this agent's feedback bought? No account, no API key

Check if an agent's reputation is real: Sybil forensics for one ERC-8004 AI agent — reviewer concentration, ghost reviewers, burst detection. Is this agent's feedback bought? No account, no API key

Decide whether to interact with an AI agent at all: Can I interact with this agent? INTERACT/CAUTION/AVOID verdict with confidence for conversations, task delegation and data copying — not just payments. Required: agent_id or solana_asset. Shareable verdict page in every response

Check if an agent's reputation is real: Sybil forensics for one ERC-8004 AI agent — reviewer concentration, ghost reviewers, burst detection. Is this agent's feedback bought? No account, no API key

Use this to check a draw independently: recomputes the SHA-256 of a revealed server_seed against its hash, and the numbers of a swoge-hmac-sha256-v1 draw, or the 52-card shoe of a SWOGE casino hand (scheme swoge-casino-shoe with its nonce). Price: $0.003 in USDC on Base or Solana, or $0.02 in USDG or $SWOGE on Robinhood Chain (tool price + settlement cost, minimum $0.02).