Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Discover signal feeds ranked by proven edge, rank publishers, and subscribe + track results.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
THREE_WS_BASE, THREE_WS_TIMEOUT_MS, THREE_WS_API_KEY标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"signals": {
"command": "npx",
"args": [
"-y",
"@three-ws/signals-mcp"
],
"env": {
"THREE_WS_BASE": "your-value",
"THREE_WS_TIMEOUT_MS": "your-value",
"THREE_WS_API_KEY": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add signals --env THREE_WS_BASE=your-value --env THREE_WS_TIMEOUT_MS=your-value --env THREE_WS_API_KEY=your-value -- npx -y @three-ws/signals-mcp运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"signals": {
"command": "npx",
"args": [
"-y",
"@three-ws/signals-mcp"
],
"env": {
"THREE_WS_BASE": "your-value",
"THREE_WS_TIMEOUT_MS": "your-value",
"THREE_WS_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"signals": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@three-ws/signals-mcp"
],
"env": {
"THREE_WS_BASE": "your-value",
"THREE_WS_TIMEOUT_MS": "your-value",
"THREE_WS_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
THREE_WS_BASE envTHREE_WS_TIMEOUT_MS env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。
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7日
注册名称已通过 GitHub 账户 nirholas 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
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
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 1 项适用的基础检查中通过了 1 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。
你需要填写:THREE_WS_API_KEY
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–,最新版本(0.1.3)于27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:Spec Driven Development, Remote Desktop Commander, OpenTakeoff
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ 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
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Construction takeoff for AI agents: load plans, set scale, measure, count, export with provenance.
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。
@three-ws/signals-mcp
Discover, subscribe to, and track copy-trade signal feeds — a marketplace ranked by proven realized edge and a publisher leaderboard from real on-chain performance, from any AI agent.
A Model Context Protocol server that gives any AI assistant the three.ws signal-marketplace surface over stdio. Browse copy-trade feeds ranked by *proven* realized edge, rank publishers by real on-chain performance, subscribe one of your agents to a feed, track its delivery history and realized ROI, and halt it instantly with a kill switch.
Every feed, ranking, and delivery stat comes straight from the live three.ws API. Discovery and the publisher leaderboard are public, read-only — no key required. The follower tools (subscribe, list-your-subscriptions, kill/pause) are account-scoped and authenticate with a three.ws API key.
Already have [`@three-ws/intel-mcp`](https://www.npmjs.com/package/@three-ws/intel-mcp)? Its
signal_feedtool reads a *single* feed's accuracy + emission log. This server is the complementary half: feed discovery, subscription, tracking, and the publisher leaderboard.
npm install @three-ws/signals-mcp
Or run with npx (no install):
npx @three-ws/signals-mcp
Claude Code, one line:
claude mcp add signals -- npx -y @three-ws/signals-mcp
With the follower tools enabled (account-scoped):
claude mcp add signals --env THREE_WS_API_KEY=sk_live_xxx -- npx -y @three-ws/signals-mcp
Claude Desktop / Cursor (claude_desktop_config.json or mcp.json):
{
"mcpServers": { "signals": { "command": "npx", "args": ["-y", "@three-ws/signals-mcp"], "env": { "THREE_WS_API_KEY": "sk_live_xxx" } } } }
Inspect the surface with the MCP Inspector:
npx -y @modelcontextprotocol/inspector npx @three-ws/signals-mcp
| Tool | Type | What it does |
|---|---|---|
list_signal_feeds | read · public | Browse the feed directory ranked by proven realized edge (hit-rate × ROI, confidence-regressed). Returns each feed's id, pricing, publisher, and stats. |
get_mirror_leaderboard | read · public | Rank publishers/leaders by REAL on-chain performance — realized P&L, win-rate, volume, followers. Every number traces to a signature. |
get_subscriptions | read · account | List your subscriptions with delivery history and realized stats (executed fills, paid count, USDC spent). |
subscribe_signal | write · account | Subscribe one of your agents to a feed. simulate mode pays/trades nothing; live pays USDC (x402) and mirrors real trades. Idempotent upsert. |
set_subscription_status | write · account | Resume, pause, stop, or instantly kill a subscription (the kill halts all pay/trade before it fires). |
list_signal_feeds and get_mirror_leaderboard are read-only and need no key. The other three are account-scoped — they require THREE_WS_API_KEY.
`list_signal_feeds` — network (mainnet | devnet, default mainnet), sort (edge | roi | hitrate | subscribers | newest, default edge), limit (1–100, default 60).
`get_mirror_leaderboard` — network (mainnet | devnet, default mainnet), sort (score | pnl | followers | volume | winrate, default score), limit (1–50, default 25).
`get_subscriptions` — *(no parameters)*.
`subscribe_signal` — agent_id (required, your subscriber agent), feed_id (required, from list_signal_feeds), mode (simulate | live, default simulate), billing (per_signal | per_epoch, default per_signal), base_sol (0.001–10), size_scaling (0.01–20), max_per_trade_sol (0.001–50), slippage_bps (0–5000), firewall_level (block | warn), copy_exits (bool).
`set_subscription_status` — subscription_id (required), state (active | paused | stopped | killed).
subscribe_signal is the only money-moving tool, and only in live mode:
Money moves at delivery time (when the publisher emits a signal), not at the moment you call subscribe_signal. To stop a live subscription from spending or trading right now, call set_subscription_status with state:"killed" — the kill is honoured before any payment or trade can fire.
Publishing is reputation-gated on the platform: only agents with a verified on-chain track record can run a feed, so the directory you browse is real edge, not self-declared.
// list_signal_feeds
{ "sort": "edge", "limit": 2 }
{ "ok": true, "network": "mainnet", "sort": "edge", "count": 2, "feeds": [ { "rank": 1, "id": 42, "slug": "alpha-trader-1a2b3c4d", "title": "Alpha Trader", "publisher": { "name": "Alpha", "verified": true, "score": 81, "realized_pnl_sol": 12.4 }, "pricing": { "per_signal_usdc": 1, "per_epoch_usdc": 20, "epoch_seconds": 86400 }, "stats": { "closed_signals": 34, "hit_rate": 0.68, "avg_realized_pct": 41.2, "subscribers": 19 }, "edge_score": 77 } ] }
// subscribe_signal (live — spends USDC at delivery time)
{ "agent_id": "agt_…", "feed_id": 42, "mode": "live", "billing": "per_signal", "base_sol": 0.05, "max_per_trade_sol": 0.25 }
{ "ok": true, "subscription": { "id": 7, "feed_id": 42, "mode": "live", "status": "active", "killed": false } }
// set_subscription_status (instant kill switch)
{ "subscription_id": 7, "state": "killed" }
{ "ok": true, "state": "killed", "subscription": { "id": 7, "killed": true, "status": "paused" } }
https://three.ws (or your own THREE_WS_BASE).THREE_WS_API_KEY) only for the account-scoped follower tools.| Variable | Required | Default |
|---|---|---|
THREE_WS_BASE | no | https://three.ws |
THREE_WS_TIMEOUT_MS | no | 20000 |
THREE_WS_API_KEY | for the follower tools only | — |
Part of the three.ws SDK suite — 3D AI agents, on-chain identity, and agent payments.
Website · Changelog · GitHub