Procore MCP Server

Procore MCP Server

Enables ChatGPT agents to read Procore project data and create confirmed daily log notes, with project allowlisting and OAuth.

Category
访问服务器

README

Procore MCP Server for ChatGPT Agents

This is a Railway-ready starter MCP server for connecting ChatGPT PM and Superintendent agents to Procore.

The server sits between ChatGPT and Procore:

ChatGPT Agent -> Custom MCP app -> Railway MCP server -> Procore API

What This Starter Includes

  • A public MCP endpoint at /mcp
  • A health check at /health
  • Procore OAuth client-credentials token handling
  • Project allowlisting with PROCORE_ALLOWED_PROJECT_IDS
  • Read tools for projects, daily log notes, RFIs, and submittals
  • A write tool for creating Daily Log notes that requires confirmed: true
  • An optional raw Procore request tool for development only

Required Railway Variables

Add these in Railway under the service's Variables tab:

PROCORE_BASE_URL=https://api.procore.com
PROCORE_CLIENT_ID=your_procore_client_id
PROCORE_CLIENT_SECRET=your_procore_client_secret
PROCORE_COMPANY_ID=your_company_id
PROCORE_ALLOWED_PROJECT_IDS=comma,separated,project,ids

Optional:

MCP_BEARER_TOKEN=make_a_long_random_value
ENABLE_RAW_PROCORE_TOOL=false

Use MCP_BEARER_TOKEN only if your ChatGPT custom MCP setup supports passing a bearer token. Otherwise leave it blank during initial testing, then add proper OAuth protection before broader production use.

Deploy To Railway

From the Railway screen you are on:

  1. Choose GitHub Repository if you have pushed this folder to GitHub.
  2. Or choose Empty Project, then create a service from GitHub later.
  3. Set the environment variables above.
  4. Let Railway run:
npm install
npm run build
npm run start
  1. Open Railway's generated domain.
  2. Confirm this works:
https://your-railway-domain.up.railway.app/health
  1. In ChatGPT's Custom MCP form, use:
https://your-railway-domain.up.railway.app/mcp

ChatGPT Custom MCP Settings

Use these values in the ChatGPT app editor:

  • Name: Procore Tools
  • Description: Lets the PM and Superintendent agents read approved Procore project data and create confirmed project records.
  • Server URL: https://your-railway-domain.up.railway.app/mcp
  • Authentication: start with No Authentication only for private testing, or use OAuth/Bearer protection if available.

Safety Rules

Do not put Procore secrets in ChatGPT instructions. Keep them only in Railway variables.

Keep PROCORE_ALLOWED_PROJECT_IDS populated. This prevents the server from writing to projects outside your approved list.

Keep ENABLE_RAW_PROCORE_TOOL=false after early development. The raw tool is intentionally not suitable for production because it can call arbitrary Procore API paths.

Write tools require confirmed: true. Your agent instructions should say:

Before using any Procore write tool, summarize the exact project, date, destination tool, and record content, then ask the user to confirm.

First Test Prompts

After connecting the MCP app to ChatGPT, test with:

List the Procore projects this tool can see.
Show me today's daily log notes for project 12345.
Create a daily log note for project 12345 for 2026-07-29 saying "Site walk completed; no safety issues observed." Ask me to confirm before submitting it.

Notes About Procore Endpoints

This starter uses common Procore REST paths for projects, daily log note logs, RFIs, and submittals. Procore tools can vary by account configuration and permissions. If an endpoint returns a validation error, use the error message and Procore API reference to adjust the request body for your enabled tools.

Useful Procore references:

  • OAuth client credentials: https://developers.procore.com/documentation/oauth-client-credentials
  • Daily Logs API: https://developers.procore.com/reference/rest/daily-logs
  • Notes Logs API: https://developers.procore.com/reference/rest/notes-logs
  • RFIs API: https://developers.procore.com/reference/rest/rfis
  • Submittals API: https://developers.procore.com/reference/rest/submittals

Useful OpenAI references:

  • Build an MCP server: https://developers.openai.com/plugins/build/mcp-server
  • Connect and test in ChatGPT: https://developers.openai.com/plugins/deploy/connect-chatgpt

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选