Strategy Voting

多策略投票交易系统。当用户说"投票策略"、"多策略"、"策略投票"、"综合信号"时触发。整合突破、RSI均值回归、MACD金叉、布林带四个策略，加权投票决定交易信号。

Security Scan

The skill's code and instructions match its stated purpose (a multi-strategy signal generator) and there are no unexplained network calls, credential requests, or installs — main issues are missing dependency notes and a hard-coded run path, not malicious behavior.

Purpose & Capability

Name/description (多策略投票交易系统) align with the included Python implementation: four strategies (breakout, RSI reversion, MACD cross, Bollinger) and weighted voting logic are implemented. The skill does not request unrelated credentials or binaries.

Instruction Scope

SKILL.md tells the agent to run the bundled script at /root/.openclaw/workspace/skills/strategy-voting/scripts/voter.py which matches the provided file. The runtime instructions do not attempt to read unrelated files or environment variables. Note: the instructions only run a self-contained test dataset — they do not explain how to feed live market data or integrate with an exchange, and they do not declare required Python packages.

Install Mechanism

There is no install spec (instruction-only), so nothing external is downloaded. The code imports numpy and pandas but the skill does not declare or install these dependencies; this is an operational omission (may cause runtime errors) rather than a security red flag. No external URLs or archives are used.

Credentials

The skill declares no required environment variables or config paths and the code does not access environment secrets or network endpoints. There are no credential requests — proportional to being a local signal generator.

Persistence & Privilege

always is false and the skill does not attempt to modify other skills or system configuration. It only prints a report and does not persist data or enable itself globally.

Assessment

What to consider before installing: (1) This skill appears to do exactly what it claims — generate weighted voting signals — and does not exfiltrate data or request credentials. (2) Operational notes: the script requires numpy and pandas but the skill provides no install steps; install those packages in the agent environment or add a dependency spec. (3) SKILL.md runs the script via a hard-coded workspace path — ensure the agent environment has the file at that location or adjust the path. (4) The included script uses a generated test dataset and has no broker/exchange integration — if you plan to connect it to live trading, review and add secure, explicit code for authentication and order execution (never paste API keys into untrusted skills). (5) If you need the agent to call this autonomously, be aware autonomous invocation is allowed by default; because this skill requests no secrets the risk is low, but always review integration code before giving it access to real funds.

多策略投票系统

综合多个技术指标策略，加权投票做出交易决策。

策略组成

策略	权重	说明
突破	1.5x	价格突破20日高点+成交量放大（回测最佳）
RSI均值回归	1.0x	RSI超卖(<30)买入，超买(>70)卖出
MACD金叉	1.0x	MACD线上穿/下穿信号线
布林带	1.0x	价格触及上下轨

使用方法

python3 /root/.openclaw/workspace/skills/strategy-voting/scripts/voter.py

决策规则

• 综合得分 ≥ 1.5 → 做多
• 综合得分 ≤ -1.5 → 平仓/做空
• 其他 → 观望

集成

可与现有交易机器人配合，作为信号过滤器。