MCP adapter for x402station.io, the independent risk-signal layer for x402 agentic commerce. Exposes Preflight by x402station.io plus Forensics, Catalog Decoys, Alternatives, Credits, Watch, and Whats New. Any agent speaking the Model Context Protocol gets endpoint evidence before signing PAYMENT-SIGNATURE — decoy, zombie, price-trap, never-paid, latency, signature/settlement checks — before paying.
x402station.io independently probes every endpoint listed on agentic.market every 10 minutes and merges probe history with CDP settlement data. Policy engines decide and enforce; x402station.io measures and reports. We do not route, take custody, or endorse.
# Claude Code / Cursor / Windsurf / Continue — works anywhere with MCP
npm install -g x402station-mcp
# or use npx in the config, no global install needed:Related MCP server: RMI MCP + x402 Gateway
The adapter charges real USDC per paid call through x402 itself. You need a wallet private key that holds Base mainnet USDC.
Add to ~/.claude/claude_desktop_config.json (or wherever your MCP servers live):
{
"mcpServers": {
"x402station": {
"command": "npx",
"args": ["-y", "x402station-mcp"],
"env": {
"AGENT_PRIVATE_KEY": "0xYOUR_PRIVATE_KEY_HERE"
}
}
}
}Same shape — every MCP host understands command / args / env. See your tool's MCP docs.
Variable | Required | Default | Purpose |
| yes for any paid tool call | — | 0x-prefixed 64-hex-char private key. Account must hold Base mainnet USDC. |
| no |
| Override for dev / testing. |
preflight(url) — $0.001 USDCAsk whether it's safe to pay this x402 URL. Returns:
{
"ok": true,
"warnings": [],
"metadata": {
"service": "...",
"price_usdc": "0.01",
"uptime_1h_pct": 100,
"avg_latency_ms": 412
}
}ok is true only when no critical warning fires. Warnings include unknown_endpoint, no_history, dead, zombie, decoy_price_extreme, never_paid_zombie, proxy_markup, wildcard_402, spa_fallback, suspicious_high_price, slow, and new_provider.
forensics(url) — $0.001 USDCDeep 7-day report. Superset of preflight. Returns hourly uptime, latency p50/p90/p99, status-code distribution, concentration-group stats, decoy probability. Extra warnings: dead_7d, mostly_dead, slow_p99, price_outlier_high, high_concentration.
catalog_decoys() — $0.005 USDCFull blacklist. Returns every active endpoint currently flagged critical, plus per-reason counts. Pull periodically and cache locally — cheaper than preflighting every URL.
agent wants data from some_endpoint.com
│
├─ preflight("https://some_endpoint.com/data") ← $0.001
│ ok: false, warnings: ["decoy_price_extreme"]
│
└─ skip; try the next candidateFor bulk discovery, do catalog_decoys() once per day and treat the result as a set-difference against any URLs you're about to hit.
x402station checks whether the endpoint is safe (decoy / zombie / price-trap / never-paid) before an agent pays. It pairs naturally with a payload safety check — screening the payment metadata for PII before signing:
agent intent: pay some_endpoint.com
├─ x402station preflight(url) ← is the ENDPOINT safe?
├─ presidio-x402 screen_payment_metadata() ← is the PAYLOAD safe?
└─ pay()presidio-hardened-x402-mcp screens payment metadata for PII before signing. The two servers are developed independently, on purpose — keeping the signals uncorrelated is the point. This is a composition pointer, not an endorsement; see their repo for the screening tool's behaviour and limits.
Service: https://x402station.io
Manifest: https://x402station.io/.well-known/x402
Dataset: https://huggingface.co/datasets/x402station/preflight-dataset-v0_1
x402 spec: https://x402.org
General + commercial: hello@x402station.io
Bug reports: https://github.com/sF1nX/x402station-mcp/issues
Security disclosures (RFC 9116): https://x402station.io/.well-known/security.txt
MIT
Given a URL flagged by preflight (or a taskClass hint), returns up to 5 healthy sibling endpoints in the same provider/domain/category/price-band. Filters out 7-day-dead and 1-hour-erroring candidates; ranks by uptime + latency. Costs $0.005 USDC. Use this immediately after preflight returns ok=false — it answers 'where do I go instead?'. Pass {url} when you have a specific URL the agent was about to pay; pass {taskClass} (e.g. 'llm-completions', 'Inference') when discovering by service category; or both for a richer match.
Inputs: url, taskClass, limit.
Bulk-prepaid preflight bundle. Pay $0.50 USDC once for 1000 prepaid /api/v1/preflight calls. Effective rate $0.0005/call (50% off the per-call $0.001 tier). Returns { creditId, balance: 1000, expiresAt }. STORE THE creditId — it's the bearer token and is not retrievable later. Pass it via X-Credit-Id header on subsequent /api/v1/preflight calls; on exhaustion (balance=0) or expiry (90 days) the middleware falls through to per-call x402 automatically. Use this once you've decided to do high-volume preflight work.
Inputs: no named parameters.
Returns every active x402 endpoint currently flagged decoy_price_extreme / zombie / dead_7d / mostly_dead in one JSON payload, plus per-reason counts. Costs $0.005 USDC. Pull periodically (hourly/daily) and cache locally as a blacklist — cheaper than preflighting every URL.
Inputs: no named parameters.
Free, no payment required. Returns { creditId, balance, initialBalance, used, paidAmount, createdAt, expiresAt, expired, paymentTx, paymentNetwork }. UUID-only access — anyone holding the creditId can read state, same as decrement. 404 covers both 'malformed UUID' and 'no such credit' (same body so an attacker scraping random UUIDs can't tell them apart).
Inputs: creditId (required).
Deep history for one x402 endpoint: hourly uptime over 7 days, latency p50/p90/p99, status-code distribution, concentration-group stats (how crowded this provider's namespace is), and a decoy probability score [0, 1]. Costs $0.001 USDC. Superset of preflight — if you're running forensics you don't need preflight too.
Inputs: url (required).
Ask x402station whether a given x402 URL is safe to pay. Returns {ok, warnings[], metadata}. Costs $0.001 USDC (auto-signed with AGENT_PRIVATE_KEY). Call this BEFORE any other paid x402 request to avoid decoys (price ≥ $1k), zombie services, and dead endpoints. ok:true only when no critical warning fires.
Inputs: url (required).
Returns the current state of a watch: active/expired, alerts remaining (out of 100 prepaid), last 10 alert deliveries with their delivery_status, and the last computed signal snapshot. Free — no payment required, secret-gated. The secret is the one returned by watch_subscribe.
Inputs: watchId (required), secret (required).
Pay $0.01 USDC for a 30-day watch + 100 prepaid alerts on one x402 endpoint. When subscribed signals fire or clear (e.g. endpoint goes zombie, price flips to decoy_price_extreme), x402station POSTs a JSON payload signed with HMAC-SHA256 to your webhookUrl. Returns watchId + secret — STORE THE SECRET, it's the HMAC seed for verifying delivery payloads and is not retrievable later. signals defaults to the critical preflight subset {dead, zombie, decoy_price_extreme}; pass other names to subscribe to non-critical signals too.
Inputs: url (required), webhookUrl (required), signals.
Deactivates the watch — no further alerts will be queued or delivered. The subscription is set is_active=false but the row + alert history is retained for audit. Free — no payment required, secret-gated. There is no refund for unused prepaid alerts.
Inputs: watchId (required), secret (required).
Polling-friendly catalog diff. Body { since?, limit? } (default since=now-24h, limit=200, max 500). Returns added_endpoints[] (first_seen_at >= since AND is_active=true), removed_endpoints[] (flipped to is_active=false since), service-level counts, polls_in_window, and current active totals. Cheap ($0.001 USDC) so hourly polling stays under $1/month — perfect for aggregator agents that need a fresh delta without re-pulling the whole catalog. Internal ingest cron runs every 5 min, so polling more often than that returns identical data.
Inputs: since, limit.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 10 tools
Each tool serves a distinct function: preflight checks, alternative endpoints, credit management, forensics, catalog blacklists, watch subscriptions, and catalog diffs. Descriptions are detailed, leaving no ambiguity about tool purposes.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., buy_credits, watch_subscribe, whats_new), making predictions and selection straightforward for an agent.
With 10 tools, the server covers its domain of endpoint safety monitoring, credit management, and subscription alerts without being overly bloated or sparse.
The tool surface covers the main workflow: preflight checks, alternatives, credit buying, status, forensics, blacklisting, watching, and polling. A minor gap is the lack of a free endpoint status query for a single URL, but the paid preflight covers it.
Preflight before paying any x402 tool (ok|caution|avoid + history). 1 free/wallet; then $0.01 Base.
Parallel Solana token risk and transaction safety for AI agents. Paid MCP tools inspect tokens and unsigned transactions; the HTTP risk panel queries compatible advisors in parallel and returns an indicative consensus. x402-compatible clients pay USDC automatically, with no signup or API key.
Pay-per-call ($0.005-$0.03 USDC) market, on-chain, and prediction-market data API for AI agents and trading bots via x402 — no signup, no API key. Remote MCP server, 28 tools: token security (honeypot/liquidity/counterfeit checks), Hyperliquid oracle-gap risk, kimchi premium, funding APR, DEX slippage, Polymarket arbitrage/liquidity audits, Hyperliquid HIP-4 odds.
Free token-safety scans + paid x402 verdicts, signals, radar & EVM swap quotes for AI agents.
