Skip to content
Get packs

Guide · Agent skill

How to use the CCXT Crypto Market Data agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The ccxt skill teaches an AI agent to fetch public cryptocurrency market data, such as historical candles (OHLCV), latest prices, order books and recent trades, from more than 100 exchanges using the open-source CCXT library. It is aimed at analysts, quants and developers who want free price data for backtests or charts without creating an exchange account. After this guide you will know what the skill does, how to install it, how to phrase requests so the agent uses it, and what limits to keep in mind.

CCXT Crypto Market DataHKUDS/Vibe-Trading · MIT

Quick start

  1. 1Download the ZIP. Download the ccxt skill ZIP from this page.
  2. 2Unzip into a skills folder. Unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder.
  3. 3Or upload in the Claude apps. Open Settings → Capabilities → Skills and upload the ZIP.
  4. 4Install the ccxt package. Run pip install ccxt in your Python environment.
  5. 5Ask for market data. Request data in plain language, naming the symbol, time range and output you want.
  6. 6Verify the result. Check candles and prices for gaps or errors before using them in analysis.

What is an agent skill, and what does the ccxt skill do?

An agent skill is a folder containing a SKILL.md instruction file that an AI agent reads when a task matches the skill's description. The skill gives the agent working knowledge of a specific tool, so it does not have to guess how to use it.

The ccxt skill covers public market data. CCXT is a unified library supporting exchanges such as Binance, Bybit, OKX, Coinbase and Kraken. With this skill the agent knows how to request historical candles, latest tickers, order books and recent trades, how to write symbols such as BTC/USDT, which timeframes to use, and how to choose an exchange.

  • Historical price candles (OHLCV): timestamp, open, high, low, close, volume
  • Latest quote for a symbol (ticker)
  • Order books and recent trades
  • Public data only: the skill does not cover trading or personal accounts

How do you install the ccxt skill?

Download the ZIP for the ccxt skill from this page. Then either unzip it into your personal skills folder, ~/.claude/skills/, or into a project's .claude/skills/ folder so it applies only to that project.

Alternatively, in the Claude apps, open Settings → Capabilities → Skills and upload the ZIP there.

The skill needs Python with the ccxt package. Install it with the command below.

  • Personal: unzip into ~/.claude/skills/
  • Per project: unzip into the project's .claude/skills/
  • Claude apps: Settings → Capabilities → Skills, then upload the ZIP
pip install ccxt

How do you trigger the skill with natural requests?

You do not need to name the skill. Ask for crypto market data in plain language and mention the asset, the quote currency and the kind of data you need. Naming the exchange is optional; the default is Binance.

Good requests state a symbol, a time range or number of candles, and the output you want, for example a chart, a table or a CSV.

  • Get the last 100 daily BTC/USDT candles and summarise the trend.
  • What is the current ETH/USDT price on OKX?
  • Pull the order book for SOL/USDT, top 20 levels per side.
  • Fetch 4-hour ETH/BTC candles for the past year for a backtest.

What does the skill teach the agent about symbols and timeframes?

CCXT uses a slash format for symbols, such as BTC/USDT, ETH/BTC or SOL/USDT. If the Vibe-Trading project's DataLoader receives a hyphenated symbol such as BTC-USDT, it converts it to BTC/USDT automatically.

Supported timeframes are 1m, 3m, 5m, 15m, 30m, 1h, 2h, 4h, 6h, 12h, 1d, 1w and 1M. Not every exchange supports every timeframe, so the agent can check the exchange's timeframes list before requesting data.

  • Symbol format: BTC/USDT (slash)
  • Timeframes: 1m to 1M, as listed above
  • Check exchange.timeframes for the exchange in use

How does the agent fetch data? A quick code example

The skill shows the agent a basic Python pattern. It creates an exchange object with rate limiting enabled, requests daily candles, and fetches a ticker. Candles come back as lists of timestamp, open, high, low, close and volume.

For long histories, CCXT pages through results using the since parameter, a millisecond timestamp. The Vibe-Trading built-in loader does this automatically, up to 200 pages.

  • fetch_ohlcv(symbol, timeframe, since, limit): historical candles
  • fetch_ticker(symbol): latest quote with last, bid, ask, volume
  • fetch_tickers(symbols): batch quotes
  • fetch_order_book(symbol, limit): bids, asks, timestamp
  • fetch_trades(symbol, since, limit): recent trades
import ccxt exchange = ccxt.binance({"enableRateLimit": True}) # Fetch daily OHLCV ohlcv = exchange.fetch_ohlcv("BTC/USDT", "1d", limit=100) # Returns: [[timestamp, open, high, low, close, volume], ...] # Fetch ticker ticker = exchange.fetch_ticker("ETH/USDT") print(f"ETH price: {ticker['last']}")

What are the safety notes and limits?

This skill runs scripts and uses the network. It needs internet access to exchange public endpoints. It needs no API key for public market data, and it is not designed for placing trades or accessing a personal account. Any trading would need separate credentials and carries financial risk.

Exchange data can have gaps, delays or errors, and availability depends on third-party exchanges and their rate limits. The skill is a fallback for the Vibe-Trading project when the OKX loader is unavailable, so it is not a guaranteed production feed. Verify important figures before making financial decisions.

  • Runs scripts: yes. Uses network: yes.
  • No API key needed for public data; do not add trading keys for this skill.
  • Check data for gaps and errors before relying on it.
  • Rate limits and exchange downtime are outside your control.

What people use it for

  • Pulling daily BTC/USDT candles to backtest a simple moving-average strategy.
  • Checking the latest ETH/USDT price on a specific exchange for a dashboard.
  • Downloading 4-hour ETH/BTC history for chart analysis.
  • Inspecting the top of the SOL/USDT order book at a point in time.
  • Providing a fallback price source for the Vibe-Trading backtesting project when the OKX loader is unavailable.

Frequently asked questions

Is the ccxt skill free to use?

The skill itself is listed under the MIT licence in its source project, and CCXT is an open-source library. Public market data needs no API key, so you can fetch it without creating an exchange account. Check the licence terms of the source project and the exchanges' own data terms for your use.

What do I need to install before using the ccxt skill?

You need Python with the ccxt package, installed with pip install ccxt, and internet access to exchange public endpoints. No API key is needed for public market data. Then place the skill ZIP in ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it in Settings → Capabilities → Skills.

Do I need an API key for CCXT market data?

No. The skill states that public market data such as OHLCV candles, tickers and order books requires no API key. Trading or account access would need separate credentials, which this skill does not cover.

Which exchanges does the ccxt skill support?

CCXT supports 100+ exchanges, including Binance, Bybit, OKX, Coinbase, Kraken, Bitget and Gate. The default in the skill is Binance, and you can set another with the CCXT_EXCHANGE environment variable. Not every exchange supports every timeframe, so check exchange.timeframes.

Why is my timeframe or symbol not working?

Use slash format such as BTC/USDT; the Vibe-Trading DataLoader converts BTC-USDT automatically. Also confirm the timeframe is in the supported list and that the exchange supports it via exchange.timeframes, since support varies by exchange.

How does ccxt compare with using an exchange's own API?

CCXT gives one consistent interface across 100+ exchanges, so the same method names work on each. The source does not compare it with individual exchange APIs in terms of speed or coverage, so test the exchange you need before relying on it.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Use this skill if you want free, keyless public crypto prices and candles for backtesting, charts or a fallback data source in the Vibe-Trading project. Pick something else if you need to trade, access a personal account, or depend on a guaranteed production data feed, since exchange availability, rate limits and data gaps are outside the skill's control.

Download the skill