trading-skills MCP — 使用場景, 安裝 & 即時演示

為什麼要用

核心特性

Real-time signals across 50+ crypto perpetual markets every minute
44 scientific methods backed by 53 academic citations
Three AI experts with consensus voting and confidence scoring (0-100%)
Complete signals: entry, stop-loss, take-profit, leverage recommendation
Performance tracking with hit rate, cumulative ROI, and profit factor

即時演示

實際使用效果

trading-skill-2.replay ▶ 就緒

0/0

安裝

選擇你的客戶端

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "trading-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/roman-rr/trading-skills",
        "~/.claude/skills/trading-skills"
      ],
      "_inferred": true
    }
  }
}

開啟 Claude Desktop → Settings → Developer → Edit Config。儲存後重啟應用。

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "trading-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/roman-rr/trading-skills",
        "~/.claude/skills/trading-skills"
      ],
      "_inferred": true
    }
  }
}

Cursor 使用與 Claude Desktop 相同的 mcpServers 格式。專案級設定優先於全域。

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "trading-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/roman-rr/trading-skills",
        "~/.claude/skills/trading-skills"
      ],
      "_inferred": true
    }
  }
}

點擊 Cline 側欄中的 MCP Servers 圖示，然後選 "Edit Configuration"。

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "trading-skill-2": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/roman-rr/trading-skills",
        "~/.claude/skills/trading-skills"
      ],
      "_inferred": true
    }
  }
}

格式與 Claude Desktop 相同。重啟 Windsurf 生效。

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "trading-skill-2",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/roman-rr/trading-skills",
        "~/.claude/skills/trading-skills"
      ]
    }
  ]
}

Continue 使用伺服器物件陣列，而非映射。

~/.config/zed/settings.json

{
  "context_servers": {
    "trading-skill-2": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/roman-rr/trading-skills",
          "~/.claude/skills/trading-skills"
        ]
      }
    }
  }
}

加入 context_servers。Zed 儲存後熱重載。

claude mcp add trading-skill-2 -- git clone https://github.com/roman-rr/trading-skills ~/.claude/skills/trading-skills

一行命令搞定。用 claude mcp list 驗證，claude mcp remove 移除。

使用場景

實戰用法： trading-skills

How to receive and analyze crypto trading signals

👤 Crypto traders who want AI-analyzed market signals ⏱ ~10 min beginner

何時使用： You want to see what trading opportunities the system has identified.

前置條件

Trading skill installed — /plugin install roman-rr/trading-skills or claude mcp add trading-signals --transport http -- https://signals.x70.ai/mcp

步驟

Check current signals

What trading signals are active right now? Show me the ones with confidence above 70%.✓ 已複製

→ List of active signals with coin, direction, confidence, entry, SL, TP
Analyze the reasoning

Explain the transmission chain for the top signal. What data points led to this recommendation?✓ 已複製

→ 2-4 step causal reasoning with specific technical indicators

結果： Understanding of current market opportunities with AI-backed analysis.

注意事項

Treating signals as guaranteed trades — Always do your own research. Signals are analytical tools, not financial advice.

Track trading signal performance and optimize strategy

👤 Traders evaluating the system's accuracy over time ⏱ ~5 min beginner

何時使用： You want to see historical performance metrics before acting on signals.

前置條件

Trading skill installed and running — Install as above

步驟

Review performance

Show me the trading signal performance for the last 30 days. What's the hit rate, cumulative ROI, and profit factor?✓ 已複製

→ Performance metrics with per-coin and aggregate statistics

結果： Data-driven understanding of signal reliability.

注意事項

Past performance doesn't guarantee future results — Use paper trading to validate before risking real capital

組合

與其他 MCP 搭配，撬動十倍槓桿

trading-skill-2 + filesystem

Log signals and performance data to files for your own analysis

Log all signals from today with their reasoning chains to ~/trading/signals-2025-04-14.json.✓ 已複製

工具

此 MCP 暴露的能力

工具	輸入參數	何時呼叫
get_signals	min_confidence?: float	Get current active trading signals
get_performance	period?: str	Get historical signal performance
get_signal_detail	signal_id: str	Get the full reasoning chain for a specific signal

成本與限制

運行它的成本

API 配額: Served from signals.x70.ai
每次呼叫 Token 數: 300–1000 tokens per signal query
費用: Free during beta — no credit card required
提示: Check signals periodically rather than every minute to conserve tokens.

安全

權限、密鑰、影響範圍

憑證儲存： N/A — no authentication required during beta

資料出站： Signal data fetched from signals.x70.ai

Trading signals are for informational purposes only — not financial advice.
Crypto markets are highly volatile. Never trade more than you can afford to lose.
This tool does NOT execute trades. You must place trades manually on your exchange.
Always use paper trading to validate signals before using real capital.

故障排查

常見錯誤與修復

No signals available

Signals are only generated when all 3 AI experts reach consensus. During quiet markets, fewer signals appear.

驗證： Try with a lower confidence threshold

Server unreachable

Verify signals.x70.ai is accessible. Check your network connection.

驗證： curl https://signals.x70.ai/mcp

Signal seems stale

Signals update every minute. If a signal is old, it may have expired. Check the timestamp.

驗證： Get fresh signals and check timestamps

替代方案

trading-skills 對比其他方案

替代方案	何時用它替代	權衡
Manual technical analysis	You prefer to do your own analysis with TradingView or similar tools	Full control but no AI automation

trading-skills

為什麼要用

核心特性

即時演示

實際使用效果

安裝

選擇你的客戶端

使用場景

實戰用法： trading-skills

How to receive and analyze crypto trading signals

前置條件

步驟

注意事項

Track trading signal performance and optimize strategy

前置條件

步驟

注意事項

組合

與其他 MCP 搭配，撬動十倍槓桿

工具

此 MCP 暴露的能力

成本與限制

運行它的成本

安全

權限、密鑰、影響範圍

故障排查

常見錯誤與修復

替代方案

trading-skills 對比其他方案

更多

資源