← 返回 Skills 市场
1132
总下载
1
收藏
6
当前安装
1
版本数
在 OpenClaw 中安装
/install crypto-backtest
功能描述
Crypto futures backtesting engine with built-in EMA, RSI, MACD, and Bollinger Band strategies. Fetches OHLCV data from any ccxt-supported exchange (Bybit, Bi...
使用说明 (SKILL.md)
Crypto Backtest Engine
Fast, scriptable backtesting for crypto futures strategies. Fetches data via ccxt, runs strategies, reports metrics.
Quick Start
pip install ccxt numpy
python scripts/backtest_engine.py --symbol ETH/USDT:USDT --strategy ema --fast 12 --slow 26
Features
- Multi-exchange: Any ccxt-supported exchange (Bybit, Binance, OKX, Bitget...)
- Built-in strategies: EMA crossover, RSI, MACD, Bollinger Bands
- Parameter sweep: Test all combinations automatically
- Risk simulation: Configurable leverage, position size, SL/TP, fees
- JSON export: Machine-readable results for pipeline integration
- Custom strategies: Simple plug-in interface
Usage
Single Strategy
python scripts/backtest_engine.py \
--symbol SOL/USDT:USDT \
--strategy rsi \
--period 14 --oversold 30 --overbought 70 \
--capital 1000 --leverage 5
Parameter Sweep
python scripts/sweep.py \
--symbol ETH/USDT:USDT \
--strategies ema,rsi,macd,bbands \
--capital 1000 --leverage 5 \
--output results.json
Custom Strategy
See references/custom_strategy.md for the plug-in interface.
Output Metrics
Each backtest reports:
- Total trades, win rate, profit factor
- Total PnL (absolute + percentage)
- Max drawdown
- Best/worst trade
- Final balance
Files
scripts/backtest_engine.py— Core engine with EMA, RSI, MACD, Bollinger Bandsscripts/sweep.py— Multi-strategy parameter sweep runnerreferences/custom_strategy.md— Guide for adding custom strategiesreferences/strategy_notes.md— Notes on each built-in strategy's edge cases
安全使用建议
This skill appears to be what it claims: a local backtesting engine that pulls public market data via ccxt, runs strategies, and writes JSON results. Before running: (1) review the included Python files locally (no obfuscation or hidden endpoints were found), (2) run pip installs in a virtual environment to avoid contaminating your system Python, (3) be aware it will make network requests to exchanges (rate limits apply), (4) do not provide exchange API keys unless you understand and intend to call private endpoints, and (5) treat backtest outputs as exploratory — validate any promising strategies with paper trading before risking real funds. Also check the included scripts for any syntax errors or truncation in your copy (the prompt excerpt was truncated in places; ensure the repository files are complete before use).
功能分析
Type: OpenClaw Skill
Name: crypto-backtest
Version: 1.0.0
The OpenClaw skill bundle provides a crypto backtesting engine. All files, including the `SKILL.md` instructions, are aligned with the stated purpose. The Python scripts (`backtest_engine.py`, `sweep.py`) use standard libraries like `ccxt` for data fetching and `argparse` for command-line arguments. There is no evidence of data exfiltration, malicious execution (e.g., `curl|bash`, `eval` of untrusted input), persistence mechanisms, or prompt injection attempts against the AI agent. File I/O is limited to saving backtest results locally, which is an expected function. The code is straightforward and lacks obfuscation or suspicious dependencies.
能力评估
Purpose & Capability
Name/description match the included code and SKILL.md. The code fetches OHLCV via ccxt and implements EMA, RSI, MACD, and Bollinger strategies and a sweep runner — all coherent with a backtest engine. No unexpected binaries or credentials are required.
Instruction Scope
SKILL.md instructs installing ccxt and numpy and running the provided scripts. The runtime instructions and the code operate only on market data fetched via ccxt and local parameter configs; there are no instructions to read unrelated system files, exfiltrate data to third-party endpoints, or access secrets.
Install Mechanism
There is no packaged install spec. The README suggests 'pip install ccxt numpy' which is a standard PyPI install for this functionality. No downloads from arbitrary URLs or archive extraction are used.
Credentials
The skill declares no required environment variables or credentials. That fits the code, which uses ccxt for public OHLCV fetches. Note: ccxt supports private APIs (requiring API keys) for trading/account endpoints — those are not needed for this skill unless the user modifies the code to call private endpoints.
Persistence & Privilege
always is false and the skill does not request persistent agent privileges. It writes only local output files (results JSON) as expected. It does not modify other skill configurations or system-wide settings.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install crypto-backtest - 安装完成后，直接呼叫该 Skill 的名称或使用
/crypto-backtest触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
- Initial release of crypto-backtest: a fast, scriptable backtesting engine for crypto futures.
- Supports fetching OHLCV data from any ccxt-supported exchange (Bybit, Binance, OKX, and more).
- Includes built-in EMA crossover, RSI, MACD, and Bollinger Band strategies.
- Enables multi-strategy parameter sweeps with automatic metric reporting (win rate, PnL, drawdown, etc.).
- Exports machine-readable backtest results to JSON for integration with trading pipelines.
- Simple plug-in interface for custom strategies.
元数据
常见问题
Crypto Backtest 是什么？
Crypto futures backtesting engine with built-in EMA, RSI, MACD, and Bollinger Band strategies. Fetches OHLCV data from any ccxt-supported exchange (Bybit, Bi... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1132 次。
如何安装 Crypto Backtest？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install crypto-backtest」即可一键安装，无需额外配置。
Crypto Backtest 是免费的吗？
是的，Crypto Backtest 完全免费（开源免费），可自由下载、安装和使用。
Crypto Backtest 支持哪些平台？
Crypto Backtest 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Crypto Backtest？
由 SunnyZhou（@sunnyztj）开发并维护，当前版本 v1.0.0。
推荐 Skills