OpenContext
Enables interaction with open data portals through a configurable plugin system, supporting CKAN and custom data sources.
README
Esri UC 2026 MCP Server
An MCP server for the 2026 Esri User Conference (San Diego, July 13–17): session catalog, expo directory, time-aware schedule ("what's on right now"), schedule-block planning with conflict detection, and room locations — built on the OpenContext one-fork-one-server framework.
Endpoint: https://esri-uc.codeforanchorage.org/mcp
Provenance
All data is a static snapshot of the RainFocus public session catalog
(event.esri.com, widget esri/26uc — the same API that powers the public
detailed agenda),
taken 2026-07-14 UTC with scripts/snapshot.py: 982 sessions, 321
exhibitors, and 223 room geometries from the official
UC Event Map's
backing ArcGIS service (the conference venue is itself a feature service).
Nothing behind attendee login is accessed — no My Schedule, reservations, or
waitlists. The server makes zero upstream calls at runtime; every response
carries the snapshot date, and a catalog refresh is simply
python scripts/snapshot.py + redeploy. Room changes happen — attendees
should verify last-minute changes in the Esri Events app.
Try these prompts
- "What MCP or agentic AI sessions are on Wednesday afternoon, and do any of them conflict?"
- "What's starting in the next hour that's relevant to local government?"
- "Where is my next session, and is there coffee within a five-minute walk
of that room?" (composes
where_is_room's lat/lng with maps tools)
Tools
| Tool | Answers |
|---|---|
search_sessions |
keyword + day/type/track/time-window search |
get_session |
full record: abstract, speakers, room, livestream |
whats_on |
in progress now / starting soon, venue-local clock |
plan_block |
candidates for a time block, overlap conflicts marked |
find_exhibitors |
expo directory with booth numbers |
get_tracks_and_types |
exact filter vocabulary (types, tracks, days…) |
where_is_room |
building, floor, lat/lng, event-map deep link |
Quick start (local)
cp config-example.yaml config.yaml # esri_uc is already enabled in config.yaml
python3 scripts/local_server.py # http://localhost:8000/mcp
Refresh the snapshot (public catalog only; sequential, 1 s delay between
requests): python scripts/snapshot.py, review the diff in data/, redeploy
with ./scripts/deploy.sh --environment prod.
Connect via Claude Connectors: Settings → Connectors → Add custom connector → paste the endpoint URL.
Framework documentation
| Doc | Description |
|---|---|
| Getting Started | Setup and usage |
| Architecture | System design and plugins |
| Deployment | AWS, Terraform, monitoring |
| Testing | Local testing (Terminal, Claude, MCP Inspector) |
Built on OpenContext by Srihari Raman (City of Boston DoIT); this fork by Code for Anchorage. MIT licensed — see LICENSE.
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。