Token导航 LogoToken导航

Agent Skill · 研究检索

clawback(clawback 搜索)

clawback 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

clawback 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 1,116。

总安装 119,160周安装 4,965GitHub Stars 公开资料未说明下载 39,720核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.1.1

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:clawback(clawback 搜索)
来源仓库:https://github.com/mainfraame/clawback
安装命令:
openclaw skills install clawback
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install clawback

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

clawback 用于根据国会股票交易报告执行自动化经纪操作。适合希望模拟合规交易策略并进行风险管理的研究者使用。通过 clawhub 安装,集成国会数据源与券商接口模拟层。使用前需验证数据来源合法性、交易规则匹配度及回测框架兼容性。建议仅用于教育或模拟环境,避免接入真实资金账户。

clawback 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install clawback;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

clawback 怎么安装?

在终端执行 openclaw skills install clawback。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

clawback 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

clawback 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

clawback 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

clawback 用的人多吗?

公开数据是累计安装 119,160 次,近一周安装 4,965 次,下载 39,720 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

clawback 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
clawback

clawback 搜索 · 本页

119,160—接触敏感数据MIT-0有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
clawback
description
Mirror congressional stock trades with automated broker execution and risk management. Use when you want to track and automatically trade based on congressional disclosures from House Clerk and Senate eFD sources.
version
1.1.0
author
mainfraame
homepage
https://github.com/mainfraame/clawback
user-invocable
true
metadata
{"openclaw": {"emoji": "🦀", "requires": {"bins": ["python3", "pip"]}, "install": {"pip": "{baseDir}"}, "primaryEnv": "BROKER_API_KEY"}}

ClawBack

Mirror congressional stock trades with automated broker execution

Agent Instructions

When the user invokes /clawback, execute the appropriate command based on the argument:

Commands

When the user invokes /clawback with any arguments, execute the corresponding command:

CommandAction
/clawback setupRun the setup wizard: Execute {baseDir}/bin/clawback.py setup
/clawback statusCheck system status: Execute {baseDir}/bin/clawback.py status
/clawback runStart trading bot: Execute {baseDir}/bin/clawback.py run
/clawback daemonRun as background service: Execute {baseDir}/bin/clawback.py daemon
/clawback testTest notifications: Execute {baseDir}/bin/clawback.py test
/clawback (no args)Show help: Execute {baseDir}/bin/clawback.py --help

How to Execute Commands

Option 1: Using the wrapper script (recommended) When executing ClawBack commands, always:

  1. Use the wrapper script at {baseDir}/bin/clawback.py
  2. Pass the command as an argument (e.g., {baseDir}/bin/clawback.py status)
  3. Capture and display the output to the user

Option 2: Direct Python execution (if wrapper doesn't work) If the wrapper script fails, you can run ClawBack directly:

  1. Change to the skill directory: cd {baseDir}
  2. Activate the virtual environment: source venv/bin/activate
  3. Run the CLI: python -m clawback.cli [command]
  4. Capture and display the output

Important: Always check if the virtual environment exists at {baseDir}/venv. If not, you may need to run the setup first.

/clawback setup - Interactive Setup Flow

When user runs /clawback setup, follow these steps:

Step 1: Install dependencies (if needed) Check if {baseDir}/venv exists. If not, run:

cd {baseDir} && python3 -m venv venv && source venv/bin/activate && pip install -e .

**Step 2: Prompt for E*TRADE credentials** Ask the user for each value:

  1. Environment: Ask "Do you want to use sandbox (testing) or production (real money)?"

- Default: sandbox

  1. Consumer Key: Ask "Enter your E*TRADE Consumer Key (from developer.etrade.com):"

- Required field

  1. Consumer Secret: Ask "Enter your E*TRADE Consumer Secret:"

- Required field

  1. Account ID: Ask "Enter your E*TRADE Account ID (or leave blank to get it after OAuth):"

- Optional - can be obtained later

Step 3: Save configuration Create/update ~/.clawback/config.json with the provided values:

{
  "broker": {
    "adapter": "etrade",
    "environment": "<sandbox or production>",
    "credentials": {
      "apiKey": "<consumer_key>",
      "apiSecret": "<consumer_secret>"
    }
  },
  "trading": {
    "accountId": "<account_id>",
    "initialCapital": 50000,
    "tradeScalePercentage": 0.01,
    "maxPositionPercentage": 0.05,
    "dailyLossLimit": 0.02
  },
  "notifications": {
    "telegram": {
      "enabled": true,
      "useOpenClaw": true
    }
  },
  "congress": {
    "dataSource": "official",
    "pollIntervalHours": 24,
    "minimumTradeSize": 10000
  }
}

Step 4: Confirm setup Tell the user: "Configuration saved to ~/.clawback/config.json. Run /clawback status to verify."

Getting E*TRADE API Credentials

Direct user to: https://developer.etrade.com

  1. Create a developer account
  2. Create a new app (sandbox first for testing)
  3. Copy the Consumer Key and Consumer Secret

Configuration Location

  • Config file: ~/.clawback/config.json
  • Skill directory: {baseDir}

Reading Saved Configuration

To check if the user has configured credentials, read ~/.clawback/config.json:

  • If file doesn't exist or credentials are empty → prompt for setup
  • If credentials exist → can proceed with status/run commands

The CLI automatically reads from ~/.clawback/config.json for all operations.

Checking Setup Status

Before running /clawback status or /clawback run, verify:

  1. {baseDir}/venv exists (dependencies installed)
  2. ~/.clawback/config.json exists with non-empty broker.credentials.apiKey

If either is missing, suggest running /clawback setup first.


ClawBack tracks stock trades disclosed by members of Congress (House and Senate) and executes scaled positions in your E*TRADE brokerage account. Built on the premise that congressional leaders consistently outperform the market due to informational advantages.

Default Target Politicians

ClawBack monitors these politicians by default (configurable):

PoliticianChamberPriority
Nancy PelosiHouse1 (highest)
Dan CrenshawHouse2
Tommy TubervilleSenate2
Marjorie Taylor GreeneHouse3

Trading Strategy Defaults

ParameterDefaultDescription
Trade Delay3 daysWait after disclosure before trading
Holding Period30 daysTarget hold time for positions
Position Size5%Max allocation per trade
Stop-Loss8%Per-position stop-loss
Portfolio Drawdown15%Max portfolio loss before halt
Disclosure Checks10:00, 14:00, 18:00 ETDaily check times

Features

  • Real-time disclosure tracking from official House Clerk and Senate eFD sources
  • Automated trade execution via E*TRADE API (only supported broker)
  • Smart position sizing - scales trades to your account size
  • Trailing stop-losses - lock in profits, limit losses
  • Risk management - drawdown limits, consecutive loss protection
  • Telegram notifications - get alerts for new trades and stop-losses
  • Backtesting engine - test strategies on historical data

Performance (Backtest Results)

StrategyWin RateReturnSharpe
3-day delay, 30-day hold42.9%+6.2%0.39
9-day delay, 90-day hold57.1%+4.7%0.22

Congressional leaders have outperformed the S&P 500 by 47% annually according to NBER research.

Installation via ClawHub

# Install from ClawHub registry
clawhub install clawback

# Or install from local directory
clawhub install ./clawback

Troubleshooting

Common Issues

  1. Skill not executing: If /clawback doesn't work in OpenClaw:

- Check if the skill is in the correct location: {baseDir}/ - Verify the wrapper script is executable: chmod +x {baseDir}/bin/clawback.py - Check if virtual environment exists: {baseDir}/venv/

  1. Authentication issues: If E*TRADE authentication fails:

- Run the authentication utility: python {baseDir}/scripts/auth_utility.py --auth - Run {baseDir}/bin/clawback.py setup to reconfigure - Check credentials in ~/.clawback/config.json - Verify E*TRADE API keys are valid

  1. Token expiration: If tokens expire (30-day lifespan):

- Run: python {baseDir}/scripts/auth_utility.py --refresh - Or start new authentication: python {baseDir}/scripts/auth_utility.py --auth

  1. Python import errors: If you see "ModuleNotFoundError":

- Ensure virtual environment is activated - Run pip install -e . in {baseDir}/ - Check Python path includes {baseDir}/src

Debug Mode

To debug skill execution, add DEBUG=1 environment variable:

DEBUG=1 {baseDir}/bin/clawback.py status

This will show additional information about the execution context.

Post-Installation Setup

After installation via ClawHub, the install.sh script runs automatically:

  1. Python Environment Setup - Creates virtual environment
  2. Package Installation - Installs ClawBack via pip
  3. Directory Structure - Creates logs/, data/, config/ directories
  4. Setup Prompt - Asks if you want to run the setup wizard

If you skip setup during installation, run it manually:

cd ~/.openclaw/skills/clawback
./setup.sh          # Interactive setup wizard
# or
clawback setup      # CLI-based setup

Improved Setup Features

  • Better input handling - Works in both interactive and non-interactive modes
  • Input validation - Validates E*TRADE API key formats
  • Timeout handling - Automatically uses defaults if no input
  • Error recovery - Fallback to manual setup if CLI fails
  • Configuration check - Detects existing config and offers options

Interactive Setup Wizard

The setup wizard guides you through configuration:

Step 1: Environment Selection

  • Sandbox (recommended for testing): No real trades, uses E*TRADE developer sandbox
  • Production: Real trading with real money

Step 2: E*TRADE API Credentials

  • Consumer Key: From E*TRADE developer portal
  • Consumer Secret: From E*TRADE developer portal

Step 3: Authentication

  • Automatic OAuth flow with E*TRADE
  • Opens browser for authorization
  • Returns verification code

Step 4: Account Selection

  • Lists all available E*TRADE accounts
  • Choose which account to trade with

Step 5: Telegram Setup (Optional)

  • Configure notifications via Telegram bot
  • Uses OpenClaw's built-in Telegram channel if available

Environment Variables

After setup, credentials are stored in .env:

# E*TRADE API (required)
BROKER_API_KEY=your_consumer_key_here
BROKER_API_SECRET=your_consumer_secret_here
BROKER_ACCOUNT_ID=your_account_id_here

# Telegram (optional)
TELEGRAM_BOT_TOKEN=your_bot_token_here
TELEGRAM_CHAT_ID=your_chat_id_here

# FMP API (optional)
FMP_API_KEY=your_fmp_api_key_here

Usage

# Use the installed CLI command
clawback run      # Start interactive trading mode
clawback daemon   # Run as background service
clawback status   # Check system status
clawback setup    # Re-run setup wizard
clawback test     # Test Telegram notifications

Automated Trading

The clawback daemon command runs continuously with:

  • Disclosure checks at 10:00, 14:00, 18:00 ET (when filings are typically released)
  • Trade execution at 9:35 AM ET (5 min after market open)
  • Token refresh every 90 minutes (keeps E*TRADE session alive)
  • Market hours enforcement (9:30 AM - 4:00 PM ET)

Data Sources

  • House Clerk: https://disclosures-clerk.house.gov (PDF parsing)
  • Senate eFD: https://efdsearch.senate.gov (web scraping)
  • Financial Modeling Prep: Enhanced financial data (optional)

Supported Brokers

ClawBack currently only supports E*TRADE. The adapter pattern allows for future broker support, but only E*TRADE is implemented and tested.

BrokerAdapterStatus
E*TRADEetrade_adapter.pySupported

Risk Management

  • Position limits: 5% max per symbol, 20 positions max
  • Stop-losses: 8% per position, 15% portfolio drawdown
  • Daily limits: 3% max daily loss
  • PDT compliance: Conservative 2 trades/day limit

Authentication Helpers

For manual E*TRADE authentication outside the main CLI:

# Standalone OAuth authentication script
cd {baseDir}
source venv/bin/activate
python scripts/auth_script.py

This generates an authorization URL, prompts for the verification code, and completes authentication.

File Locations

FilePurpose
~/.clawback/config.jsonMain configuration
~/.clawback/.access_tokens.jsonE*TRADE OAuth tokens
~/.clawback/data/trading.dbSQLite database

Security

  • No hardcoded credentials in source code
  • Environment variable based configuration
  • Encrypted token storage for E*TRADE
  • Git-ignored .env file
  • Optional production encryption

Support

  • Documentation: See README.md for detailed setup
  • Issues: https://github.com/mainfraame/clawback/issues
  • Community: https://discord.com/invite/clawd

Disclaimer

Trading involves substantial risk of loss. This software is for educational purposes only. Past congressional trading performance does not guarantee future results. Always test with E*TRADE sandbox accounts before live trading.

平台分布

OpenClaw

83.82%
按下载量换算33,293

安全审计

VirusTotal

可疑

ClawScan

可疑

Static analysis

未展示

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills