Oracle EPM Cloud MCP Server

Oracle EPM Cloud MCP Server

Connect AI agents to Oracle EPM Cloud (Planning, PBCS, EPBCS) via REST APIs. Automate month-end close, run business rules, manage substitution variables, refresh databases, export data slices, and monitor job status through natural language. Built by Fred Mamadjanov, Oracle ACE and EPM Solution Architect.

Category
访问服务器

README

Oracle EPM Cloud MCP Server

Connect Claude AI (or any MCP-compatible client) to Oracle EPM Cloud via REST APIs.

Built by Fred Mamadjanov — Oracle ACE, EPM Solution Architect | fmepm.com


What This Does

This MCP server gives AI agents the ability to interact with Oracle EPM Cloud. Instead of manually running Postman calls or EPM Automate commands, you can ask Claude to:

  • "What applications are in my EPM environment?"
  • "Show me the current substitution variables"
  • "Run the Agg_AllData business rule"
  • "Export Q1 revenue data for North America"
  • "Roll the current month forward from Mar to Apr"

The server translates these natural language requests into Oracle EPM REST API calls.


Architecture

You (natural language) → Claude Desktop → MCP Protocol → This Server → Oracle EPM REST APIs → Your EPM Cloud

Available Tools

Tool What It Does EPM REST API
get_api_version Test connectivity, discover API versions GET /HyperionPlanning/rest/
list_applications List all EPM applications GET /HyperionPlanning/rest/v3/applications
get_substitution_variables Read current month, year, scenario variables GET .../substitutionvariables
run_business_rule Execute a calc script or business rule POST .../jobs
check_job_status Poll whether a job completed or errored GET .../jobs/{jobId}
export_data_slice Pull data from the cube by dimension members POST .../exportdataslice
update_substitution_variable Change a substitution variable value PUT .../substitutionvariables

Quick Start (Mock Mode — No EPM Environment Needed)

Prerequisites

Step 1: Download and install

# Clone or download this folder
cd oracle-epm-mcp-server
npm install

Step 2: Configure Claude Desktop

Open Claude Desktop → Settings → Developer → Edit Config

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "oracle-epm-cloud": {
      "command": "node",
      "args": ["/FULL/PATH/TO/oracle-epm-mcp-server/index.js"]
    }
  }
}

Replace /FULL/PATH/TO/ with the actual path on your machine.

Windows example:

{
  "mcpServers": {
    "oracle-epm-cloud": {
      "command": "node",
      "args": ["C:\\Users\\Fred\\oracle-epm-mcp-server\\index.js"]
    }
  }
}

Mac example:

{
  "mcpServers": {
    "oracle-epm-cloud": {
      "command": "node",
      "args": ["/Users/fred/oracle-epm-mcp-server/index.js"]
    }
  }
}

Step 3: Restart Claude Desktop

Quit Claude Desktop completely and reopen it. You should see the MCP tools icon (hammer) in the chat input area. Click it to verify "oracle-epm-cloud" is listed.

Step 4: Try it

Type in Claude Desktop:

"What EPM applications are available in my environment?"

Claude will use the list_applications tool and return the mock data.


Switching to Live Mode (Real EPM Environment)

When you have access to an Oracle EPM Cloud environment, set these environment variables in your Claude Desktop config:

{
  "mcpServers": {
    "oracle-epm-cloud": {
      "command": "node",
      "args": ["/FULL/PATH/TO/oracle-epm-mcp-server/index.js"],
      "env": {
        "EPM_MODE": "live",
        "EPM_BASE_URL": "https://epm-YOURDOMAIN.epm.REGION.oraclecloud.com",
        "EPM_USERNAME": "IDENTITYDOMAIN.your_username",
        "EPM_PASSWORD": "your_password",
        "EPM_APP_NAME": "Vision"
      }
    }
  }
}

That's it. Same 7 tools, but now hitting your real environment.

Authentication Note

The REST APIs use Basic Authentication. Your username format must be identitydomain.username — this is the #1 most common mistake. If you get 401 errors, check this first.

Accounts with multi-factor authentication (MFA) enabled cannot use Basic Auth. You would need OAuth 2.0 instead (not covered in this version).


Month-End Close Automation Example

Here's the sequence an AI agent would follow to automate a month-end close:

  1. Check current period: get_substitution_variables → sees CurrMonth = "Mar"
  2. Run aggregation: run_business_rule with "Agg_AllData"
  3. Wait for completion: check_job_status with the returned job ID
  4. Validate data: export_data_slice for Revenue, COGS, Net Income
  5. Roll period forward: update_substitution_variable CurrMonth from "Mar" to "Apr"
  6. Confirm: get_substitution_variables → verifies CurrMonth = "Apr"

This is the exact same workflow a finance team does manually each month — now executable through natural language.


Troubleshooting

Error Cause Fix
MCP tools not showing in Claude Config path wrong Check claude_desktop_config.json path is absolute
401 Unauthorized Username format Use identitydomain.username, not just username
403 Forbidden Insufficient permissions User needs EPM admin or appropriate role
Connection refused Wrong URL Check EPM_BASE_URL matches your environment
ETIMEDOUT Network/firewall Verify you can reach the EPM URL from your machine

What's Next

  • OAuth 2.0 support — for environments with MFA enabled
  • FCCS-specific tools — consolidation, intercompany elimination
  • Data integration tools — file upload/download via Migration APIs
  • Groovy rule execution — run Groovy scripts via REST API

About

This server was built as part of the demonstration of connecting an AI agent to Oracle EPM Cloud. For more Oracle EPM content, tutorials, and tools:


This is not an Oracle product. Oracle EPM Cloud is a trademark of Oracle Corporation.

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
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
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
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
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
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
mcp-server-qdrant

mcp-server-qdrant

这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。

官方
精选
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