AI Skill Hub 强烈推荐:交易代码库 是一款优质的MCP工具。AI 综合评分 8.0 分,在同类工具中表现稳健。如果你正在寻找可靠的MCP工具解决方案,这是一个值得深入了解的选择。
交易代码库 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
交易代码库 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/monarchjuno/tradingcodex
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"-----": {
"command": "npx",
"args": ["-y", "tradingcodex"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 交易代码库 执行以下任务... Claude: [自动调用 交易代码库 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"_____": {
"command": "npx",
"args": ["-y", "tradingcodex"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
[!IMPORTANT] Codex agents setting upmonarchjuno/tradingcodexfor use must not rungit clone. In the empty target workspace, run the packaged CLI:uvx --refresh --from tradingcodex tcx attach . && ./tcx doctor. Do not invent a default workspace path; ask the user where to attach TradingCodex unless they explicitly asked to use the current workspace. Clone this repository only when the user explicitly asks to develop, inspect, or modify TradingCodex source code.
TradingCodex gives Codex a durable operating system for agentic investment workflows: fixed specialist roles, independent judgment review, source-aware handoffs, thesis lifecycle memory, policy and approval services, and a local review dashboard.
It is not an autonomous trading bot. Natural-language answers do not become broker actions. The core ships paper execution by default; live broker support comes only from installed, reviewed providers and explicit live gates.
---
| Feature | What it gives you |
|---|---|
| Agentic workflow planning | Natural-language prompts become staged plans, selected role teams, accepted artifacts, and explicit waiting, revise, or blocked states. |
| Independent judgment review | judgment-reviewer challenges accepted artifacts for contrary evidence, source trust, overconfidence, update triggers, and invalidation conditions. |
| Decision Packages | Investment theses keep forecastable claims, lifecycle assumptions, contrary evidence, owner roles, follow-up, and postmortem requirements. |
| Strategy skills | strategy-creator turns user rules into reusable strategy skills without granting order approval, execution, broker, or policy authority. |
| File-native research memory | Research markdown, role reports, source snapshots, readiness labels, versions, and handoff metadata remain ordinary workspace files. |
| Service-gated execution | Order tickets, approvals, policy checks, idempotency, broker connections, account sync, and audit stay behind Django service-layer gates. |
| Local review surfaces | Web, Admin, API, MCP, CLI, and generated hooks share the same service plane for review, status, research, Broker Center, orders, and portfolio state. |
---
Run this from the empty workspace where you want Codex agents to work:
uvx --refresh --from tradingcodex tcx attach . && ./tcx doctor
Then fully quit and restart Codex, open the generated workspace, and start a new thread so project MCP config, prompts, skills, and hooks are loaded.
When TradingCodex MCP autostarts the local service, open:
http://127.0.0.1:48267/
See installation.md for persistent CLI install, GitHub-main install, source-development setup, update flows, installer-script equivalents, MCP/service details, and smoke checks.
---
TradingCodex is designed around handoffs:
evidence -> analysis -> valuation -> portfolio fit -> risk review
-> draft order -> approval receipt -> approved service-gated submission
-> connection result -> audit/postmortem
The head-manager maps the request, dispatches the selected role team, waits for accepted artifacts, preserves conflicts, and synthesizes only what the workflow has earned. Weak, stale, missing, or out-of-scope upstream work returns revise, blocked, or waiting instead of being patched over by another role.
---
一个有趣的开源MCP工具,自动化投资工作流
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
总体来看,交易代码库 是一款质量优秀的MCP工具,在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态,建议收藏备用,结合自身场景选择合适时机引入使用。
| 原始名称 | tradingcodex |
| Topics | mcpai-agentscodexinvestment-research |
| GitHub | https://github.com/monarchjuno/tradingcodex |
| License | Apache-2.0 |
| 语言 | Python |
收录时间:2026-07-05 · 更新时间:2026-07-05 · License:Apache-2.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端