📦 Crypto Social Intel — 技能工具

v1.0.0

Crypto social intelligence skill for AI agents. Activate when user mentions social trends, sentiment analysis, KOL signals, mention surge, Fear & Greed Index...

0· 24·0 当前·0 累计
jackxun520 头像by @jackxun520 (Jackxun520)·MIT-0
下载技能包
License
MIT-0
最后更新
2026/4/16
0
安全扫描
VirusTotal
无害
查看报告
OpenClaw
可疑
medium confidence
The skill's stated functionality (crypto social signals via Santiment + alternative.me) is coherent, but it instructs the agent to interact with an external MCP endpoint (https://mcp-skills.ai.antalpha.com/mcp) without specifying what user data is sent or who controls that service — this creates a potential privacy/exfiltration risk that isn't documented.
评估建议
This skill appears to do what it says (crypto social signals from Santiment + alternative.me), but it relies on a third-party MCP endpoint (https://mcp-skills.ai.antalpha.com/mcp). Before installing or invoking it: 1) Ask who operates that MCP endpoint and request their privacy/data-retention policy; confirm TLS, ownership, and whether they log full user prompts or responses. 2) Do not send secrets or private data through the skill — the documentation does not specify what is forwarded. 3) If yo...
详细分析 ▾
用途与能力
The name, description, tools, and data sources (Santiment + alternative.me) align with the declared functionality (social trending, sentiment, KOL proxy, mention surges, Fear & Greed). There are no unrelated binaries or environment variables requested in the skill metadata. The note that Santiment requires an API key is consistent with the data source.
指令范围
The SKILL.md instructs the agent to initialize and call a remote MCP endpoint (https://mcp-skills.ai.antalpha.com/mcp) to run the five tools. It does not document what payloads are sent, what user context or prompts are forwarded, or the MCP operator's data handling/retention policies. Because the agent will transmit user queries and context to that external service, this is a potential privacy/exfiltration vector not accounted for in the description.
安装机制
Instruction-only skill with no install spec and no code files — nothing is written to disk or auto-installed. This is lower risk from an install/execution perspective.
凭证需求
The skill declares no required environment variables for the agent, but the SKILL.md repeatedly states that Santiment API keys are required for the Santiment-backed tools and suggests setting SANTIMENT_API_KEY on the server. That mismatch is explainable (keys may be held by the remote MCP service), but it should be explicit who controls the key and whether user-provided secrets would be uploaded to that service. No unrelated credentials are requested.
持久化与权限
The skill does not request always:true, does not request persistent system config, and is user-invocable only. It does allow normal autonomous model invocation, which is standard for skills.
安全有层次,运行前请审查代码。

License

MIT-0

可自由使用、修改和再分发,无需署名。

运行时依赖

无特殊依赖

版本

latestv1.0.02026/4/16

Initial release of crypto-social-intel skill: - Provides 5 social intelligence tools for crypto: trending tokens, sentiment score, KOL signals, mention surge detection, and real-time Fear & Greed Index. - Integrates Santiment (free tier, ~35-day lag) for most social data, and alternative.me for real-time Fear & Greed data. - Includes recommended workflows for market sentiment, token analysis, early warning, and signal combination. - Presents example response templates and clear agent behavior rules for various user queries. - Notes limitations of Santiment free tier and current KOL signals (not real Twitter data).

无害

安装命令

点击复制
官方npx clawhub@latest install crypto-social-intel
镜像加速npx clawhub@latest install crypto-social-intel --registry https://cn.longxiaskill.com

技能文档

Social intelligence layer for crypto tokens. Powered by Santiment GraphQL API + alternative.me Fear & Greed Index.

5 Tools:

  • crypto-social-trending — Top tokens by social volume
  • crypto-sentiment-score — Sentiment score for a specific token
  • crypto-kol-signals — Social dominance anomaly (KOL proxy, v1.0: social dominance as proxy, not real Twitter KOL data)
  • crypto-mention-surge — Detect abnormal mention spikes
  • crypto-fear-greed — Fear & Greed Index (real-time, free)

MCP Endpoint

https://mcp-skills.ai.antalpha.com/mcp
Environment switching: For local dev/test, replace with http://localhost:3000/mcp.

Protocol: MCP Streamable HTTP (JSON-RPC over HTTP with mcp-session-id header).

Connection Flow

1. POST /mcp → initialize (get mcp-session-id from response header)
  • POST /mcp → tools/call (with mcp-session-id header)

Data Sources

ToolSourceReal-timeAPI Key
crypto-social-trendingSantiment⚠️ Free tier: ~35-day lagRequired
crypto-sentiment-scoreSantiment⚠️ Free tier: ~35-day lagRequired
crypto-kol-signalsSantiment⚠️ Free tier: ~35-day lagRequired
crypto-mention-surgeSantiment⚠️ Free tier: ~35-day lagRequired
crypto-fear-greedalternative.me✅ Real-timeNone
Note: Santiment free tier has ~35-day data lag. For real-time social data, upgrade to Santiment Basic ($49/mo) and update SANTIMENT_API_KEY on the server — no code changes needed.

MCP Tools (5)

crypto-social-trending

Get top crypto tokens ranked by social volume.

Parameters:

  • limit (optional): 1-50, default 10
  • time_range (optional): "24h" | "7d", default "7d"

Response example:

{
  "items": [
    {
      "rank": 1,
      "slug": "bitcoin",
      "symbol": "BTC",
      "social_volume": 2413,
      "sentiment_score": 57,
      "signal_level": "LOW",
      "trend_change": "N/A",
      "data_source": "santiment"
    }
  ],
  "meta": { "data_source": "santiment", "note": "Free tier: ~35-day lag" }
}

crypto-sentiment-score

Get sentiment score and trend for a specific token.

Parameters:

  • symbol (required): Token symbol or slug, e.g. "BTC", "ETH", "bitcoin"
  • time_range (optional): "7d" | "30d", default "7d"

Response example:

{
  "symbol": "BTC",
  "sentiment_score": 57,
  "social_volume": 19301,
  "trend_direction": "down",
  "change_vs_prev": "-17.4%",
  "signal_level": "LOW",
  "data_source": "santiment"
}

crypto-kol-signals

Detect social dominance anomalies as KOL activity proxy.

v1.0 uses social_dominance_total as KOL proxy. Real Twitter KOL data planned for v2.0.

Parameters:

  • symbol (required): Token symbol or slug
  • time_range (optional): "7d" | "30d", default "7d"
  • threshold (optional): Surge multiplier threshold, 1-10, default 1.5

Response example:

{
  "symbol": "BTC",
  "social_dominance": 1.16,
  "dominance_change": -33,
  "surge_ratio": 0.67,
  "sentiment_direction": "bullish",
  "signal_level": "LOW",
  "note": "Social dominance anomaly used as KOL proxy. Real KOL data (Twitter API) planned for v2.0."
}

crypto-mention-surge

Detect tokens with abnormal social mention spikes.

Parameters:

  • threshold (optional): Surge ratio vs historical avg, default 2.0
  • time_window (optional): "7d" | "30d", default "7d"
  • limit (optional): 1-50, default 10

Response example:

{
  "items": [
    {
      "rank": 1,
      "symbol": "SOL",
      "current_volume": 1200,
      "historical_avg": 400,
      "surge_ratio": 3.0,
      "sentiment_direction": "bullish",
      "signal_level": "HIGH"
    }
  ],
  "total_found": 1
}

crypto-fear-greed

Get the Crypto Fear & Greed Index. Real-time, no API key required.

Parameters:

  • days (optional): History days to return, 1-30, default 7

Response example:

{
  "current": {
    "value": 23,
    "classification": "Extreme Fear",
    "date": "2026-04-15"
  },
  "trend": "improving",
  "signal_level": "MEDIUM",
  "market_note": "市场偏恐慌,情绪面有支撑,注意底部信号。",
  "history": [...],
  "data_source": "alternative.me"
}

Contrarian signal logic:

  • Extreme Fear (≤20) → 🔴 HIGH (historical buy opportunity)
  • Fear (21-40) → 🟡 MEDIUM
  • Neutral (41-60) → 🟢 LOW
  • Greed (61-79) → 🟢 LOW
  • Extreme Greed (≥80) → 🔴 HIGH (caution, potential top)

Signal Levels

LevelConditionBadge
HIGHsentiment>70 + surge>3x, OR Extreme Fear/Greed🔴
MEDIUMsentiment≥50 + surge≥2x, OR Fear🟡
LOWotherwise🟢
Note: crypto-kol-signals signal_level is based on dominance_change (independent of the sentiment+surge rule above). Extreme dominance change (>50% or <-50%) → HIGH; moderate change → MEDIUM; otherwise LOW.

Workflow

Check Market Sentiment (most common)

1. crypto-fear-greed { days: 7 }           ← overall market mood
  • crypto-social-trending { limit: 10 } ← what's hot
  • Present combined view to user

Analyze Specific Token

1. crypto-sentiment-score { symbol: "ETH" }
  • crypto-kol-signals { symbol: "ETH", threshold: 1.5 }
  • Combine: sentiment score + dominance signal → final assessment

Early Warning Scan

1. crypto-mention-surge { threshold: 2.0, limit: 10 }
  • For tokens with HIGH/MEDIUM signals → crypto-sentiment-score to confirm
  • Alert user to tokens with multiple converging signals

Combined with Smart Money

When social signal + on-chain signal converge:

1. crypto-mention-surge detects abnormal spike
  • → Call smart-money-signal to check if whales are buying
  • Two signals converging → stronger conviction

Message Template

When presenting social intel to user:

📊 市场情绪总览
恐惧贪婪指数: 23 — Extreme Fear 🔴
趋势: improving ↗
注记: 历史上极度恐慌区间往往是逆向机会,但需结合链上数据确认。

🔥 社交热榜 Top 5 #1 BTC vol=2413 sentiment=57 🟢 LOW #2 ETH vol=491 sentiment=52 🟢 LOW #3 SOL vol=386 sentiment=52 🟢 LOW

Mention surge alert:

🚨 社交提及暴增预警
SOL: 当前3000 vs 均值400 → 7.5x ↑ 🔴 HIGH (bullish)
建议: 关注链上是否有聪明钱跟进

Agent Behavior Rules

On "市场情绪怎么样" / "market sentiment"

  • Call crypto-fear-greed first (real-time, fast)
  • Optionally call crypto-social-trending for top tokens
  • Present combined view
  • When Santiment data is involved, always append: ⚠️ 社交数据来自 Santiment 免费层,存在约 35 天延迟,仅供参考。

On "BTC情绪" / "SOL sentiment" / specific token

  • Call crypto-sentiment-score { symbol }
  • If signal_level is HIGH or MEDIUM, also call crypto-kol-signals
  • Present score + trend + signal
  • Always append disclaimer: ⚠️ 数据来自 Santiment 免费层,存在约 35 天延迟,仅供参考,勿作实时决策依据。
  • For kol-signals output, always note: 注:v1.0 KOL 信号以社交主导度代理,非真实 Twitter KOL 数据,v2.0 将接入 Twitter API。

On "哪些币在暴涨社交" / "mention surge" / "异常热度"

  • Call crypto-mention-surge { threshold: 2.0 }
  • For HIGH signal items, offer to do deeper analysis
  • Always append disclaimer: ⚠️ 数据来自 Santiment 免费层,存在约 35 天延迟,仅供参考。

On "恐惧贪婪" / "fear greed" / "市场贪婪指数"

  • Directly call crypto-fear-greed
  • Present value, classification, trend, market note

Supported Token Slugs

Common symbol → slug mapping (Santiment):

SymbolSlug
BTCbitcoin
ETHethereum
SOLsolana
BNBbinance-coin
XRPripple
DOGEdogecoin
ADAcardano
AVAXavalanche
DOTpolkadot
LINKchainlink
UNIuniswap
ARBarbitrum
OPoptimism
SUIsui
For unlisted tokens, pass the full Santiment slug directly (e.g. "pepe", "floki").


由 Antalpha AI 提供聚合服务 | Powered by Antalpha AI

数据来源ClawHub ↗ · 中文优化:龙虾技能库