Tickory MCP Server
Scheduled scans across all Binance spot and perpetual pairs using CEL rules (RSI, volume, MAs, price action). Runs server-side 24/7, fires webhooks on match, with delivery proof and alert explainability.
README
Tickory MCP Server
Real-time crypto scanner alerts, delivered straight to your AI agent.
Tickory monitors Binance Futures markets using programmable CEL rules and sends alerts when conditions match. This MCP server lets any agent framework create scans, read alert events, and understand why alerts triggered — all through the Model Context Protocol.
Quick demo
You: "Create a scan that fires when RSI drops below 30 on any coin with volume over $100k"
Agent (via tickory_create_scan): Done — scan "Oversold Bounce" created.
You: "Run it now"
Agent (via tickory_run_scan): 3 matches found: ETHUSDT, SOLUSDT, DOGEUSDT.
You: "Why did ETH match?"
Agent (via tickory_explain_alert_event): RSI-14 was 24.7, below your threshold of 30.
Volume gate passed: $487k USDT > $100k minimum. CEL expression evaluated true.
Install
npm / npx
npx @tickory/mcp
The npm package downloads the matching GitHub Release binary for macOS and Linux during install, so npx @tickory/mcp works without a separate build step.
Go install
go install github.com/tickory/tickory-mcp@latest
Pre-built binaries
Download from GitHub Releases for Linux and macOS (amd64/arm64).
Build from source
git clone https://github.com/tickory/tickory-mcp.git
cd tickory-mcp
go build -o tickory-mcp .
Setup
- Get a Tickory account at tickory.app
- Create a scoped API key in your dashboard under Settings > API Keys
- Configure the MCP server:
export TICKORY_API_BASE_URL=https://api.tickory.app
export TICKORY_API_KEY=tk_xxxxxxxx_yyyyyyyyyyyyyyyyyyyyyyyy
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"tickory": {
"command": "npx",
"args": ["@tickory/mcp"],
"env": {
"TICKORY_API_BASE_URL": "https://api.tickory.app",
"TICKORY_API_KEY": "tk_xxxxxxxx_yyyyyyyyyyyyyyyyyyyyyyyy"
}
}
}
}
Claude Code
Add to your .mcp.json:
{
"mcpServers": {
"tickory": {
"command": "npx",
"args": ["@tickory/mcp"],
"env": {
"TICKORY_API_BASE_URL": "https://api.tickory.app",
"TICKORY_API_KEY": "tk_xxxxxxxx_yyyyyyyyyyyyyyyyyyyyyyyy"
}
}
}
}
Cursor / Windsurf / other MCP clients
Point either to npx @tickory/mcp or to the standalone tickory-mcp binary with the environment variables above. The server uses stdio (newline-delimited JSON-RPC 2.0).
Tools
| Tool | Description |
|---|---|
tickory_list_scans |
List scans visible to the API key owner |
tickory_get_scan |
Fetch one scan by ID |
tickory_create_scan |
Create a new scan with CEL expression and hard gates |
tickory_update_scan |
Replace an existing scan definition |
tickory_run_scan |
Trigger a scan run immediately |
tickory_describe_indicators |
Describe available CEL variables, recommended guards, and example expressions |
tickory_list_alert_events |
List alert events with cursor pagination |
tickory_get_alert_event |
Fetch one alert event by UUID |
tickory_explain_alert_event |
Explain why an alert triggered or was suppressed |
All tools return schema_version: "v1" for contract stability.
Configuration
| Variable | Required | Default | Description |
|---|---|---|---|
TICKORY_API_BASE_URL |
Yes | — | Tickory API base URL |
TICKORY_API_KEY |
Yes | — | Scoped API key (tk_...) |
TICKORY_TIMEOUT_SECONDS |
No | 15 |
HTTP timeout for API requests |
CLI flags (--api-base-url, --api-key, --timeout) override environment variables.
Error handling
Upstream HTTP errors are mapped to deterministic MCP error codes:
| HTTP Status | MCP Code | Retryable |
|---|---|---|
| 400 | invalid_request |
No |
| 401 | unauthorized |
No |
| 403 | forbidden |
No |
| 404 | not_found |
No |
| 409 | conflict |
No |
| 429 | rate_limited |
Yes |
| 5xx | upstream_unavailable |
Yes |
Scoped API key permissions
| Scope | What it allows |
|---|---|
read_events |
Read alert events, scan runs, activity |
manage_scans |
Create/read/update/delete scans, execute scans |
manage_routing |
Create/manage alert sources and routes |
Create keys with the minimum scopes needed. See the developer docs for details.
Protocol versions
This server negotiates MCP protocol versions: 2024-11-05, 2025-03-26, 2025-06-18, 2025-11-05, and 2025-11-25.
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。