Skip to content

skill-hub.md

Latest commit

 

History

History
155 lines (109 loc) · 7.96 KB

skill-hub.md

File metadata and controls

155 lines (109 loc) · 7.96 KB

Skill Hub

Skill Hub is a curated collection of plug-and-play AI skills bundled with Bitget Agent Hub. Each skill is a structured prompt that instructs Claude Code (or any compatible AI) how to use the market-data MCP server to answer a specific category of questions — turning raw API calls into coherent, analyst-grade outputs.

Skills live in packages/bitget-skill-hub/skills/, one subdirectory per skill, with a standard SKILL.md entry point.

What Is a Skill?

A skill is a SKILL.md file with a YAML front matter description (used for trigger matching) and a body that defines:

  • What data to fetch — which MCP tools to call and in what order
  • How to interpret it — signal thresholds, frameworks, decision trees
  • How to present it — output templates with consistent formatting
  • Error handling — neutral language when data is unavailable

Skills are self-contained. Install with npm install -g bitget-skill-hub and they activate automatically in Claude Code.

Bundled Skills

macro-analyst

Path: packages/bitget-skill-hub/skills/macro-analyst/SKILL.md

Macro-economic and cross-asset analysis for crypto market context.

What it does:

  • Reads the full rate environment: Fed funds, Treasury yield curve, 10Y–2Y spread
  • Pulls key economic indicators: CPI, Core PCE, NFP, unemployment, GDP
  • Computes BTC's rolling correlation to Gold, DXY, Nasdaq, S&P 500, VIX, 10Y yield
  • Monitors global market prices: DXY, S&P 500, Nasdaq, Gold, Oil, VIX
  • Tracks Chinese and Asian market context (Shanghai Composite, CNY/JPY/HKD)
  • Delivers a RISK-ON / MIXED / RISK-OFF verdict for crypto

Trigger phrases: macro outlook, Fed policy, FOMC, rate cut, yield curve, recession risk, inflation, CPI, DXY, cross-asset correlation, China market, global markets…

market-intel

Path: packages/bitget-skill-hub/skills/market-intel/SKILL.md

On-chain and institutional market intelligence — the structural layer beneath price.

What it does:

  • Tracks Bitcoin and Ethereum ETF flows and institutional accumulation/distribution trends
  • Surfaces whale activity: large transfers, exchange inflows/outflows, spot net flows
  • Reads market cycle indicators: AHR999, Pi Cycle, Rainbow Chart, Coinbase Premium, Bubble Index, Stablecoin supply
  • Monitors DeFi structure: TVL rankings by protocol and chain, yield pools, stablecoin market caps, protocol fees
  • Identifies trending DEX tokens and new launches (with paid-promotion disclosure)
  • Reports network health: ETH gas, BTC fees, mempool congestion

Trigger phrases: whale activity, exchange flows, ETF inflows, DeFi TVL, AHR999, rainbow chart, market cycle, DCA timing, meme coins, new token launches, gas fees…

news-briefing

Path: packages/bitget-skill-hub/skills/news-briefing/SKILL.md

Crypto and financial news aggregation, briefing, and narrative synthesis.

What it does:

  • Aggregates headlines from 20+ crypto, macro, tech, and geopolitical RSS feeds
  • Delivers morning briefings combining top stories, macro/policy news, analyst views, and community pulse
  • Supports keyword-filtered topic searches across all feeds
  • Monitors Chinese social media pulse: Weibo, Douyin, Bilibili trending topics (filtered for market-relevant content)
  • Tracks KOL and research views from key crypto thinkers
  • Synthesizes the dominant market narrative of the day

Trigger phrases: latest news, morning briefing, crypto news, what's happening, market update, Weibo trending, Chinese social media, KOL views, breaking news…

sentiment-analyst

Path: packages/bitget-skill-hub/skills/sentiment-analyst/SKILL.md

Crypto market sentiment and trader positioning analysis.

What it does:

  • Reads the Fear & Greed Index with 14-day trend context
  • Analyzes long/short ratios — both retail and top-trader accounts — to detect crowded positioning
  • Monitors funding rates to identify over-leveraged market conditions
  • Tracks taker buy/sell ratio for real-time buying/selling pressure
  • Surfaces open interest trends to gauge leverage build-up or unwinding
  • Scans Reddit community buzz for trending assets
  • Synthesizes divergences between retail and smart-money positioning

Trigger phrases: fear and greed, long/short ratio, funding rate, open interest, overleveraged, short squeeze, market mood, community sentiment, Reddit trending…

technical-analysis

Path: packages/bitget-skill-hub/skills/technical-analysis/SKILL.md

23 crypto technical indicators across 6 categories — local Python-based computation with Bitget API data.

What it does:

  • Computes 23 technical indicators across 6 categories: Trend (MA, EMA, SAR, AVL, MACD, DMI, SuperTrend), Volatility (BOLL, ATR), Oscillator (KDJ, RSI, ROC, CCI, WR, StochRSI), Volume (VOL, OBV, MFI, VWAP), Momentum (DMA, MTM, EMV), Support/Resistance (FIB)
  • Outputs time-series data (not just single points) — AI can observe trend evolution, convergence/divergence, and write richer analysis
  • Fetches kline data directly from Bitget API (spot and futures) or reads local CSV/Parquet/JSON files
  • Supports all timeframes from scalp (1min) to position (1w)
  • Provides scenario-based defaults for common queries (comprehensive analysis, trend direction, overbought/oversold, volume analysis, momentum, support/resistance, volatility)
  • Exports full indicator data to CSV for plotting or downstream backtesting

Trigger phrases: technical analysis, indicator, trend, overbought, oversold, support, resistance, momentum, volatility, volume, MACD, RSI, KDJ, BOLL, SuperTrend, EMA, MA, ATR, VWAP, FIB, StochRSI, MFI, DMI, OBV, 4h analysis, should I buy/sell…

Requirements: Python with pandas and numpy (pip install pandas numpy)

Using Skills Together

The five skills are designed to complement each other for a complete market picture:

Question type Skills to combine
Full market assessment macro-analyst + sentiment-analyst + technical-analysis
Is now a good time to buy BTC? market-intel (cycle position) + sentiment-analyst (positioning) + technical-analysis (entry levels)
What's moving markets today? news-briefing + macro-analyst
Is this altcoin worth looking at? market-intel (DEX/on-chain) + technical-analysis (TA) + sentiment-analyst (crowd positioning)

Directory Structure

packages/bitget-skill-hub/skills/
├── macro-analyst/
│   └── SKILL.md              # Macro & cross-asset analysis
├── market-intel/
│   └── SKILL.md              # On-chain & institutional intelligence
├── news-briefing/
│   └── SKILL.md              # News aggregation & narrative synthesis
├── sentiment-analyst/
│   └── SKILL.md              # Sentiment & positioning analysis
└── technical-analysis/
    ├── SKILL.md              # Technical analysis — 23 indicators, 6 categories
    ├── references/
    │   ├── indicators.md     # Indicator quick reference
    │   └── scenarios.md      # Scenario-based indicator selection
    └── src/
        ├── kline_indicators.py       # 23 indicator implementations
        └── kline_indicator_utils.py  # Utility layer & IndicatorManager

Required MCP Server

Most skills depend on the market-data MCP server (https://datahub.noxiaohao.com/mcp) for live data. Configure it in your Claude Code MCP settings before using these skills.

The server provides: crypto_market, defi_analytics, dex_market, sentiment_index, derivatives_sentiment, news_feed, social_trending, tradfi_news, network_status, rates_yields, macro_indicators, cross_asset, global_assets, global_data, cn_market.

Note: The technical-analysis skill does not use the MCP server. It fetches kline data directly from the Bitget API and runs indicator calculations locally in Python. Requires pip install pandas numpy.