World Monitor
Real-time global intelligence on conflict, markets, energy, trade, shipping and cyber. 100 tools.
Before your agent pays an x402 API, check if vet402 bought it with its own money and what came back.
列表按使用分(满分 100)排序,由以下部分组成
| 热度 | 0/80 | GitHub 星标或每周下载量:较大的计 60,较小的计 20 |
|---|---|---|
| 新鲜度 | 12/12 | 30 天内有发布或代码更新得满分 |
| 仍在响应 | 0/8 | 上次检查时云端端点有响应 |
| 使用分 | 12/100 |
质量: 基础检查通过 3/3 项(仅展示,不用于排序) · 上升势头: 暂无数据 (只用于“上升中”标签)
选择你使用的 AI 应用,然后复制设置。
用一句话让 AI 帮你安装
请帮我安装 vet402-check。按照 https://mcp.popnx.com/i/kzmttkc-vet402-check.md 操作,先把命令给我看,等我同意后再运行。粘贴到能在你电脑上工作的 AI(如 Claude Code、Cursor)。它会读取我们的步骤,先给你看命令,等你同意。无需密钥,无需注册。 · 查看 AI 会读取的全文
通过 3/3 项检查 · 在你的电脑上运行 · ✓ 命令已检查 2026年11月13日
第一步:你用哪个 AI?
选择你用的应用,步骤会显示在下方。
可连接 Claude,但需要先准备
需要先在电脑上安装 Node.js(只需一次) · 仅限电脑上的 Claude 应用(Windows 或 Mac),网页和手机不可用
把这段话复制到 Claude 的对话里。AI 会先检查电脑,告诉你要改哪个文件,并且每次都等你同意。
请在这台电脑上为 Claude 安装 MCP server vet402-check,按顺序:1) 用 node -v 检查是否有 Node.js。没有的话,请让我从 https://nodejs.org/en/download 安装,不要自己安装。2) 只使用这份配置(已通过 PopNx MCP 检查):{ "mcpServers": { "vet402-check": { "command": "npx", "args": [ "-y", "@vet402/check", "--mcp" ] } } } 3) 告诉我要修改的文件完整路径或要运行的命令,等我同意。4) 不需要密钥。5) 按约定执行,然后确认 vet402-check 的工具真的出现了。这段话不包含你的密钥。如果需要密钥,AI 会请你自己粘贴到文件或应用的输入框里。
已经装好了吗?先检查 · 打开终端,输入这条命令后按 Enter
node -vMac:按 ⌘ + Space,输入 Terminal 后按 Enter
Windows:打开“开始”菜单,输入 Windows PowerShell,然后选“打开”
还没安装: 打开官方下载页,选择 LTS 版本和点选式安装程序:Mac 用 .pkg · Windows 用 .msi。双击后一路下一步即可,不需要 curl 命令
装好后关掉终端再打开,重新运行同一条命令,应显示版本号: node -v
安装一次,所有需要 Node.js 的工具都能用
SettingsDeveloperEdit Config
{
"mcpServers": {
"vet402-check": {
"command": "npx",
"args": [
"-y",
"@vet402/check",
"--mcp"
]
}
}
}用 vet402-check 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude 说正在使用 vet402-check,或回答中出现 vet402-check 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
ChatGPT 暂时无法连接 vet402-check
vet402-check 需要安装在你的电脑上,而 ChatGPT 只能连接在网上运行的服务器
A · 使用能连接 vet402-check 的应用
B · 其他 类中能连接 ChatGPT 的其他服务器
从同类中使用较多的里挑选,仅限用地址即可连接、无需密钥的
可连接 Claude Code,但需要先准备
需要先在电脑上安装 Node.js(只需一次)
把这段话复制到 Claude Code 的对话里。AI 会先检查电脑,告诉你要改哪个文件,并且每次都等你同意。
请在这台电脑上为 Claude Code 安装 MCP server vet402-check,按顺序:1) 用 node -v 检查是否有 Node.js。没有的话,请让我从 https://nodejs.org/en/download 安装,不要自己安装。2) 只使用这份配置(已通过 PopNx MCP 检查):claude mcp add --scope user vet402-check -- npx -y @vet402/check --mcp 3) 告诉我要修改的文件完整路径或要运行的命令,等我同意。4) 不需要密钥。5) 按约定执行,然后确认 vet402-check 的工具真的出现了。这段话不包含你的密钥。如果需要密钥,AI 会请你自己粘贴到文件或应用的输入框里。
已经装好了吗?先检查 · 打开终端,输入这条命令后按 Enter
node -vMac:按 ⌘ + Space,输入 Terminal 后按 Enter
Windows:打开“开始”菜单,输入 Windows PowerShell,然后选“打开”
还没安装: 打开官方下载页,选择 LTS 版本和点选式安装程序:Mac 用 .pkg · Windows 用 .msi。双击后一路下一步即可,不需要 curl 命令
装好后关掉终端再打开,重新运行同一条命令,应显示版本号: node -v
安装一次,所有需要 Node.js 的工具都能用
Mac:按 Command+空格,输入 Terminal · Windows:打开 PowerShell
claude mcp add --scope user vet402-check -- npx -y @vet402/check --mcp应能看到 vet402-check 在列表中并显示已连接
用 vet402-check 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude Code 说正在使用 vet402-check,或回答中出现 vet402-check 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
可连接 Cursor,但需要先准备
需要先在电脑上安装 Node.js(只需一次)
把这段话复制到 Cursor 的对话里。AI 会先检查电脑,告诉你要改哪个文件,并且每次都等你同意。
请在这台电脑上为 Cursor 安装 MCP server vet402-check,按顺序:1) 用 node -v 检查是否有 Node.js。没有的话,请让我从 https://nodejs.org/en/download 安装,不要自己安装。2) 只使用这份配置(已通过 PopNx MCP 检查):{ "mcpServers": { "vet402-check": { "command": "npx", "args": [ "-y", "@vet402/check", "--mcp" ] } } } 3) 告诉我要修改的文件完整路径或要运行的命令,等我同意。4) 不需要密钥。5) 按约定执行,然后确认 vet402-check 的工具真的出现了。这段话不包含你的密钥。如果需要密钥,AI 会请你自己粘贴到文件或应用的输入框里。
已经装好了吗?先检查 · 打开终端,输入这条命令后按 Enter
node -vMac:按 ⌘ + Space,输入 Terminal 后按 Enter
Windows:打开“开始”菜单,输入 Windows PowerShell,然后选“打开”
还没安装: 打开官方下载页,选择 LTS 版本和点选式安装程序:Mac 用 .pkg · Windows 用 .msi。双击后一路下一步即可,不需要 curl 命令
装好后关掉终端再打开,重新运行同一条命令,应显示版本号: node -v
安装一次,所有需要 Node.js 的工具都能用
用 vet402-check 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Cursor 说正在使用 vet402-check,或回答中出现 vet402-check 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Cursor 帮助页面核实 · 你的界面可能因版本略有不同
可连接 VS Code,但需要先准备
需要先在电脑上安装 Node.js(只需一次)
把这段话复制到 VS Code 的对话里。AI 会先检查电脑,告诉你要改哪个文件,并且每次都等你同意。
请在这台电脑上为 VS Code 安装 MCP server vet402-check,按顺序:1) 用 node -v 检查是否有 Node.js。没有的话,请让我从 https://nodejs.org/en/download 安装,不要自己安装。2) 只使用这份配置(已通过 PopNx MCP 检查):{ "servers": { "vet402-check": { "type": "stdio", "command": "npx", "args": [ "-y", "@vet402/check", "--mcp" ] } } } 3) 告诉我要修改的文件完整路径或要运行的命令,等我同意。4) 不需要密钥。5) 按约定执行,然后确认 vet402-check 的工具真的出现了。这段话不包含你的密钥。如果需要密钥,AI 会请你自己粘贴到文件或应用的输入框里。
已经装好了吗?先检查 · 打开终端,输入这条命令后按 Enter
node -vMac:按 ⌘ + Space,输入 Terminal 后按 Enter
Windows:打开“开始”菜单,输入 Windows PowerShell,然后选“打开”
还没安装: 打开官方下载页,选择 LTS 版本和点选式安装程序:Mac 用 .pkg · Windows 用 .msi。双击后一路下一步即可,不需要 curl 命令
装好后关掉终端再打开,重新运行同一条命令,应显示版本号: node -v
安装一次,所有需要 Node.js 的工具都能用
用 vet402-check 告诉我你能帮我做什么,然后做一件给我看✓ 如果 VS Code 说正在使用 vet402-check,或回答中出现 vet402-check 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 VS Code 帮助页面核实 · 你的界面可能因版本略有不同
Claude Code
claude mcp add --scope user vet402-check -- npx -y @vet402/check --mcpClaude Desktop (claude_desktop_config.json)
{
"mcpServers": {
"vet402-check": {
"command": "npx",
"args": [
"-y",
"@vet402/check",
"--mcp"
]
}
}
}Cursor (~/.cursor/mcp.json)
{
"mcpServers": {
"vet402-check": {
"command": "npx",
"args": [
"-y",
"@vet402/check",
"--mcp"
]
}
}
}VS Code (mcp.json)
{
"servers": {
"vet402-check": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@vet402/check",
"--mcp"
]
}
}
}这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。
在你的电脑上运行,适用于
ClaudeGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatRaycast
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月19日
注册名称已通过 GitHub 账户 kzmttkc 验证,该账户拥有此仓库
代码最后更新于前天
使用 MIT
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
OSV.dev 中没有 @vet402/check@0.2.0 的记录
只检查这个版本的包本身,不包括所有依赖 · 检查于 2026年11月15日
如果你是作者,可把这段代码贴到 README。徽章会链接回本页。
[](https://mcp.popnx.com/en/servers/kzmttkc-vet402-check)提问、分享技巧或报告无法使用。按“有帮助”数量排序。
还没有评价
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 3 项适用的基础检查中通过了 3 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于前天,最新版本(0.2.0)于41天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。
同一分类中有:World Monitor, openstatus, Oh My Posh Validator, Strata, Jitsu
同一分类、其他作者、命令已通过检查。官方的排在前面,再按星标排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Real-time global intelligence on conflict, markets, energy, trade, shipping and cyber. 100 tools.
Manage monitors, status pages, incidents and maintenances in your openstatus workspace.
Validate oh-my-posh configurations and segment snippets against the official schema.
MCP server for progressive tool usage at any scale (see https://klavis.ai)
Manage Jitsu data pipelines: destinations, streams, connections, functions, live events.
| 本项 vet402-check | World Monitor | openstatus | Oh My Posh Validator | Strata | Jitsu | |
|---|---|---|---|---|---|---|
| 运行位置 | 需要安装在电脑上 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 |
| 需要密钥 | 否 | 否 | 否 | 否 | 否 | 否 |
| 可在 Claude 中按 + 添加 | 否 | 否 | 否 | 否 | 否 | 否 |
| 基础检查 | 3/3 | 4/4 | 4/4 | 4/4 | 1/4 | 4/4 |
| 许可证 | MIT | AGPL-3.0 | AGPL-3.0 | MIT | Apache-2.0 | MIT |
| 价格 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 |
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Look an x402 or MPP seller up in vet402's public record before an agent pays it.
This package reads the public record of vet402's own purchases (the vet402-delivery repository). It is separate from @vet402/mcp-server, the MCP tools for vet402's scoring API.
vet402 buys from x402 and MPP sellers with its own money on Solana, Tempo and Base (and Algorand, on its own page), checks each payment on chain, and publishes what came back: a ranking (rank.json) and signed records whose daily Merkle root is written into a Solana memo. vet402-check reads that record. It returns facts: how many purchases vet402 tried from the seller, how many settled, how many came back with an answer, the newest purchase and its tx, the failures counted against the seller and the ones that are not, the grade as rank.json prints it (measuring while there are too few purchases), and the signed records for that seller. When vet402 has never bought from the seller, the answer is "vet402 has no record of this seller". What to do with the facts is up to the caller.
Reads only public files: rank.json and records/ .json on the public site, and data/records/index.json in this repository. Verifying a record also reads public RPC. No key, no wallet, no payment, no write.
Node 22 or newer. The first run downloads the package from npm (@vet402/check); later runs start at once.
npx -y @vet402/check https://api.xona-agent.com/token/pumpfun-trending
Verify the newest signed record of that seller as well (signature, Merkle proof, payment on chain, Solana memo anchor):
npx -y @vet402/check https://api.xona-agent.com/token/pumpfun-trending --verify
As an MCP server in Claude Code (run the line above once first, so the install is cached before the client starts the server):
claude mcp add vet402-check -- npx -y @vet402/check --mcp
The same lookup over HTTP, free and with no key: GET https://vet402-delivery.vercel.app/v1/check?url= (optional chain and payTo). The answer starts with verdict and why (what the verdict means).
examples/python/check_before_paying.py. A command, a pay_after_check(url, pay) function, and a hook for the x402 Python HTTP client (x402HTTPClientSync(client).on_payment_required(vet402_on_payment_required)) that looks up the requested URL on every 402, before the client signs.vet402-check <url> [--chain solana|tempo|base|algorand|<CAIP-2>] [--pay-to <address>] [--amount <atomic units>] [--asset <address>] [--explain] [--verify] [--offline] [--json]
vet402-check verify [--offline] [--json] vet402-check diagnose --url [--chain solana|base] [--status ] [--json] vet402-check --mcp
--chain limits the answer to one chain (eip155:8453 is Base, eip155:4217 is Tempo).--pay-to compares the payTo in the 402 you hold with the payTo addresses vet402 recorded for that seller.--amount and --asset compare the 402's amount (atomic units, as x402 writes it) and asset with what vet402 paid that seller (see Compare the 402 with what vet402 paid).--explain lists the purchases the answer rests on, newest first, each with its date, chain, result, HTTP status and tx, and the payments in the seller's signed records (amount, asset, payTo).--json prints the whole result: sellers[] (one per page the seller is on), records, payTo, asOf, sources.From a clone: npm ci, then npm run check-before-paying -- .
Three read-only tools (stdio, newline-delimited JSON-RPC):
| Tool | Input | Returns |
|---|---|---|
check_before_paying | url, optional chain, pay_to, amount, asset, explain, verify_newest_record | the facts above and the named reasons, as text and as structuredContent |
verify_record | record (id or https URL), optional offline | OK, OK_NOT_ANCHORED, OK_OFFLINE or FAIL, one line per check |
diagnose_failed_payment | tx, url, optional chain, status | seller_side, facilitator_side, buyer_side or undetermined, with reasons, and an evidence pack for seller_side |
Claude Desktop (claude_desktop_config.json):
{ "mcpServers": { "vet402-check": { "command": "npx", "args": ["-y", "@vet402/check", "--mcp"] } } }
Wrap the fetch an x402 client pays through. Every 402 is looked up before the client signs:
import { wrapFetchWithPayment } from "@x402/fetch";
import { wrapFetchWithCheck } from "@vet402/check"; // npm install @vet402/check
const fetchWithPay = wrapFetchWithPayment( wrapFetchWithCheck(fetch, { onCheck: (e) => { console.log(e.checks.map((c) => c.summary).join("\n")); // The rule is yours. To stop the payment, throw here: // if (e.checks.some((c) => c.sellers.some((s) => s.settled > 0 && s.delivered === 0))) throw new Error("not paying"); }, }), client, );
The hook goes under the payment wrapper because @x402/fetch calls the fetch it was given and creates the payment only after that returns 402. The hook reads the 402 through a clone (x402 accepts from the body or the PAYMENT-REQUIRED header, and an MPP WWW-Authenticate: Payment challenge), runs check_before_paying for each chain and payTo it offers, passes the result to onCheck, and returns the 402 unchanged. It never sees a key or a signer and changes no header. A request that already carries a payment (X-PAYMENT, PAYMENT-SIGNATURE, Authorization: Payment) is the client's retry and passes straight through. If vet402's record cannot be read, onCheck gets error set and no checks.
rank.json and the records index are read once and kept in memory for 10 minutes (new PublicData({ ttlMs })), so a busy agent does not download them for every 402. The signed records read to compare amounts are kept the same way.
Every answer carries reasons[], each with one sentence and the purchases or records it rests on. They are facts next to the verdict and never change it.
| Reason | When |
|---|---|
paid_not_delivered | vet402's newest paid purchase from the seller settled and came back with no answer (rank.json's seller-side rules after a settled payment). Said only for a seller vet402 has told about its results, the same condition the verdict uses. |
payto_differs | The payTo in the 402 is not one vet402 paid this seller (and vet402 recorded at least one). |
price_jump | The 402 asks more than 10 times the highest amount vet402 paid for the same URL in the same asset. 10 times exactly is not a price_jump. |
asset_unseen | The 402 asks for an asset vet402 never paid this seller in on that chain. |
never_bought | vet402 has no purchase from this seller (on the asked chain). |
stale | vet402's newest purchase from this seller is more than 7 days old. |
price_jump and asset_unseen come from the seller's signed records (the newest 8 on that chain, those for the same URL first). With no record to compare with, neither is said.
Give the hook a policy to decide per reason:
const fetchWithPay = wrapFetchWithPayment(
wrapFetchWithCheck(fetch, { policy: { paid_not_delivered: "block", payto_differs: "block", price_jump: "block", asset_unseen: "ask_human", never_bought: "warn" }, askHuman: async (e) => confirmWithPerson(e.reasons), // true pays, anything else stops }), client, );
block throws a CheckBlockedError before the client signs; its reasons names what stopped it.ask_human calls askHuman(event) and pays only when it returns true. With no askHuman, it stops with a CheckNeedsApprovalError (a CheckBlockedError).warn and allow go on. A reason the policy leaves out is allow.onCheck gets event.reasons and event.decision (the action per reason and the strictest one) before any of this.policy, the hook decides and waits exactly as in 0.1.2: event.reasons holds the reasons that come from rank.json alone, and no signed record is read, so price_jump and asset_unseen are not looked for.event.error is set, there are no reasons, and the payment goes on, as in 0.1.2.npx -y @vet402/check https://api.xona-agent.com/token/pumpfun-trending --chain solana --amount 1000000 --asset EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v
The answer adds comparison: the signed records read, the assets vet402 paid this seller in, the highest amount vet402 paid for this URL with its record and tx, and the ratio of the 402's amount to it. With a policy, the hook makes the same comparison for every x402 offer in a 402.
npx -y @vet402/check diagnose <tx signature or hash> --url <seller URL> --status 500
From TypeScript: diagnose({ tx, url, status }, new PublicData()). As an MCP tool: diagnose_failed_payment.
SOLANA_RPC_URL and BASE_RPC_URL point it at other RPCs): settled or not, the payTo, the amount, the asset, the time. Nothing is signed or sent.The answer is fault with reasons[]:
seller_side only when all of these hold: the payment settled on chain, to a payTo vet402 paid this seller, in an asset vet402 paid it in and not below the least vet402 paid; the HTTP status is one vet402's rules put on the seller after a settled payment (5xx, 402 again); rank.json's list of the seller's newest purchases reaches back past the start of the window; vet402's own paid purchases from the seller within 24 hours also settled and came back with no answer, with none delivered; and no signed record of those days shows a delivered purchase from the seller.buyer_side only for a payment that failed on chain because the paying token account lacked the funds.facilitator_side only for a payment that failed on chain for another reason while vet402's own payments to the seller in the window also failed before settling.undetermined for everything else: the RPC could not be read, no such transaction, no status given, a 4xx, a payTo vet402 never paid, an asset vet402 never paid the seller in, an amount below the least vet402 paid, vet402 got an answer from the seller in the window (in rank.json or in a signed record of those days), rank.json's list does not reach back to the window, or vet402 has no purchase in the window. A seller is not named on less than the full test.For seller_side, evidence holds a pack to hand to the seller, as JSON and as Markdown: the tx and its explorer link, payer, payTo, amount and asset, the HTTP status the buyer reported, vet402's rule said of this payment, vet402's purchases in the window with their tx, and the signed records of those days (the payment itself left out when it is one of vet402's own). It is printed, not sent.
examples/ in this folder (not in the npm package): Coinbase AgentKit,