humanitix-mcp
A read-only MCP server for the Humanitix Public API that exposes nine tools to list events, fetch event details, list orders/tickets, check-in counts, and sales summaries.
README
humanitix-mcp
A read-only MCP (Model Context Protocol) server for the Humanitix Public API.
This repository contains the completed Phase 1-4 implementation: a uv
project, an async read-only Humanitix HTTP client, Pydantic response models,
deterministic trimmed formatting helpers, and a stdio MCP server that exposes
nine read-only tools. Phase 5 live smoke testing is pending a real API key.
Setup
This project uses uv and requires Python 3.11+.
uv sync --extra dev
Copy .env.example to .env and add a real API key when available:
cp .env.example .env
The server reads HUMANITIX_API_KEY and HUMANITIX_BASE_URL from the
environment. HUMANITIX_BASE_URL defaults to https://api.humanitix.com.
Running the server
The server communicates over stdio and is the MCP transport expected by Claude Desktop and MCP Inspector:
uv run humanitix-mcp
With the MCP Inspector:
npx @modelcontextprotocol/inspector uv run humanitix-mcp
The server creates one HumanitixClient per stdio connection and closes it
when the connection ends.
Project structure
src/humanitix_mcp/— Python package.server.py— MCP server entry point and tool handlers (Phase 4).client.py— Async Humanitix API HTTP client with auth, retry, and pagination (Phase 2).models.py— Pydantic response models (Phase 3).formatting.py— Deterministic response trimming helpers with a raw-payload opt-in escape hatch (Phase 3).
tests/— Test suite.tests/fixtures/— Redacted JSON fixtures for model and formatting tests.
Tool catalog
The server exposes nine read-only tools:
| Tool | Purpose |
|---|---|
humanitix_list_events |
List events with optional future-only and since filters. |
humanitix_get_event |
Fetch one event by event_id. |
humanitix_list_event_dates |
List the dates/sessions for an event; event-date IDs are required by the order, ticket, and check-in tools. |
humanitix_list_orders |
List orders for a specific event date. |
humanitix_get_order |
Fetch one order by order_id. |
humanitix_list_tickets |
List tickets for a specific event date, optionally filtered by status. |
humanitix_get_ticket |
Fetch one ticket by ticket_id. |
humanitix_get_check_in_count |
Get check-in totals for an event date (BETA endpoint). |
humanitix_sales_summary |
Aggregate tickets for an event date into sold/cancelled counts, gross revenue (when price data is present), and capacity utilisation (when capacity is present). |
All list tools support page_size (1-100) and max_items (default 500). All
data-returning tools support raw=true to return the full Humanitix payload
instead of the trimmed summary.
Read-only boundary
The server only implements read-only tools. No write endpoints, check-in/check-out actions, transfers, or network listeners are exposed.
Structured errors
Expected failures are returned as JSON dictionaries with success: false and a
human-readable error message. This covers missing configuration, invalid
arguments, authentication, authorization, not-found, validation, server, and
transport failures. Error messages never include the API key or raw tracebacks.
Running tests
uv run pytest
Implementation status
- [x] Phase 1: Scaffold
- [x] Phase 2: HTTP client (
client.py) - [x] Phase 3: Response models and trimming (
models.py,formatting.py) - [x] Phase 4: MCP server and tools (
server.py) - [ ] Phase 5: Live smoke test (requires a real
HUMANITIX_API_KEYand is not part of this phase)
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。