# intoApex/ai-trading-tools [Health: Active]

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/intoApex/ai-trading-tools/tree/main/mcp-server  
**GitHub Stars:** 0  
**npm Downloads (last month):** 1403  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/intoapex-ai-trading-tools

## Description
Read-only search of a sourced directory of broker MCP servers, trading APIs and market-data tools, with paper/live, order-approval and safety details.

## Tools
Capabilities this server exposes over MCP:

- **search_tools** — Search the TradingAgentStack directory of AI trading tools (broker MCP servers, trading and data APIs, bots, frameworks) by name, market or feature. Optional filters: category slug, can_trade_live, has_paper.
- **get_tool** — Full facts for one tool: pricing, live vs paper, order approval, rate limits, safety score inputs and every source URL.
- **list_category** — List every tool in a category. Call with an unknown category to get the list of categories.
- **compare_tools** — Compare two tools side by side, field by field, with sources.
- **broker_order_gating** — Who places the order for every official broker MCP server: AI trades directly, you confirm each order, approval optional, or you place the order.
- **get_risk** — Check a tool before your agent connects to it: how much money could the agent lose if it is wrong, hacked or prompt-injected? Returns the Agent Risk Score (custody, limits enforced outside the model, human approval, withdrawal scope, preview, open source, incident record), each input with its official source; missing inputs are listed, not guessed. Omit tool to rank every scored tool in a category (agent_wallet, agent_payments, ai_trading_agent, agent_token, bot_exchange, defi_agent).
- **get_incidents** — Security incidents involving AI agents, trading tools and the infrastructure they run on (breaches, exploits, security releases), each from an official disclosure. Filter by date, or by a listed tool or a bot, exchange or wallet tracked by check.
- **check** — Check before connecting (TAS Check): give a name, link, Telegram or X @handle, GitHub repo or app-store link of a Telegram trading bot, web trading terminal, exchange, swap service or wallet (incl. Zcash/Monero wallets). Returns whether the link is on that service's official list (confirmed from two independent places), a known fake, or a lookalike that uses its name, plus the worst case for your money, who holds the keys, freeze/hold policy, what a leaked API key can do, Monero/Zcash support, what the service collects about you and any data breaches, incidents, deadlines and a 3-step safe setup, each with sources. Unknown links are reported as unknown, never as safe.
- **check_setup** — Check a whole setup at once: up to 8 bots, exchanges, wallets, swap services or AI trading connectors someone uses (names, links, @handles or slugs). For each: who holds the keys, the worst case, who places orders, the Agent Risk Score or safety score, incidents in the last 12 months and the first fix, each from the dataset. Also returns the single biggest risk by a fixed rule, items we don't track (reported as unknown, never as safe), and a report_url that opens the same report on tradingagentstack.com. Links that use a tracked name but aren't official are flagged.
- **get_leaderboard** — Verified performance leaderboard for AI trading agents, bots and vaults: return, BTC over the same dates, max drawdown and Sortino ratio, computed by TradingAgentStack from the venue's public API (never screenshots, self-reported PnL or backtests), with security rated separately as a gate. Sections: ai_vaults (Hyperliquid vaults whose leader says they are AI-run, with the quoted claim), vaults (biggest Hyperliquid vaults of any strategy), agents (AI agents with an officially published trading account or on-chain registration), apps (AI trading apps on Hyperliquid: users, volume and users' realized PnL from public builder fills), models (which AI model forecasts and trades best: ForecastBench, Prophet Arena, Alpha Arena, FutureEval), signals (grade B: MQL5 robot signals on real accounts, venue-reported, not ranked), usage (fees users pay Telegram bots, trading apps and AI agent protocols, and open-source repo health). Ranked vaults passed a 30% drawdown gate and are sorted by Sortino; the rest are listed as high_drawdown. Not investment advice.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "ai-trading-tools": {
    "command": "npx",
    "args": ["-y","tradingagentstack-mcp"],
    "env": {
      "TRADINGAGENTSTACK_LIVE": ""
    }
  }
}
```

**Requires environment variables:** `TRADINGAGENTSTACK_LIVE` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What intoApex/ai-trading-tools does

intoApex/ai-trading-tools provides MCP clients with read-only access to the TradingAgentStack directory. The directory catalogs AI trading agents, broker-focused MCP servers, market-data APIs, Telegram bots, agent wallets, and exchanges. Each catalog record can include whether paper or live trading is supported, how orders are approved, a published safety score, pricing, rate limits, an official source, and when the information was last checked.

The server is intended for discovery and comparison rather than execution. It does not place trades, authenticate with brokers, or connect an agent to a brokerage account. Its data comes from the open `tools.json` dataset and is licensed separately from the server code under CC BY 4.0, with attribution to TradingAgentStack required.

## How it works

MCP clients communicate with the server over standard input/output. A client can invoke one of four read-only operations:

- `search_tools(query, category?, limit?)` finds directory entries by keyword, optionally restricting results to a category and limiting the result count.
- `get_tool(slug)` returns the full record for one directory entry, such as `alpaca-mcp-server`.
- `compare_brokers(a, b)` presents two tools for side-by-side comparison.
- `list_categories()` returns the available categories and their counts.

The npm-based setup uses the dataset bundled with the package. Setting `TRADINGAGENTSTACK_LIVE=1` makes the server try to retrieve the latest dataset from GitHub; if that request fails, it uses the bundled copy instead.

## Setup and configuration

Node.js 18 or newer is required. For clients that support npm-based MCP commands, the server can be launched with:

```bash
npx -y tradingagentstack-mcp
```

In Claude Desktop, add a server entry with `tradingagentstack` as the name, `npx` as the command, and `-y tradingagentstack-mcp` as its arguments. Cursor uses the same command in either the user-level `~/.cursor/mcp.json` file or a project-level `.cursor/mcp.json` file. Claude Code can register it with:

```bash
claude mcp add tradingagentstack -- npx -y tradingagentstack-mcp
```

Running from source requires cloning the repository, changing into `mcp-server`, installing dependencies, and running the test command. The test process copies the dataset and performs a stdio smoke test. The source entry point can then be started with `node src/index.js` or configured directly as the MCP client command.

## Tools and capabilities

The server is useful for questions such as finding broker MCP servers that support paper trading or comparing Alpaca MCP and Tradier MCP by safety controls. Search results can be narrowed by category, while individual records expose the directory’s structured details. Category counts help an agent understand the available sections before searching.

Because the server only reads the catalog, it is suitable for research, tool selection, and due diligence workflows where the agent should not be able to submit orders. It describes listed capabilities and safety information but does not validate, execute, or manage any trading activity.

## Limitations and notes

The server is not a broker connector and cannot place trades. Its results depend on the contents and freshness of the TradingAgentStack dataset. Live-data mode has a bundled-data fallback, so a failed GitHub fetch does not prevent the server from starting with packaged records.

The code is released under the MIT license. The bundled directory data uses CC BY 4.0 and should credit TradingAgentStack at `tradingagentstack.com`. The project is not financial advice.

_Full upstream README: https://allmcps.com/mcp/intoapex-ai-trading-tools/readme_