FCCS MCP Agentic Server

You are deploying an MCP Server that provides an agentic interface to FCCS (Oracle EPM Cloud Financial Consolidation and Close) with MCP support. It combines an MCP server (for Claude Desktop) and a web API to run tools, track feedback, and persist results, enabling development, testing, and integration with FCCS data workflows.

How to use

You interact with the FCCS MCP server by running the MCP server locally and then using an MCP client to issue tool calls through the web API. The server exposes endpoints to discover available tools, execute them, submit feedback, and view metrics. Start the MCP server to enable Claude Desktop integration, and use the web API to perform actions such as listing tools, running jobs, exporting data, and generating reports.

How to install

Prerequisites: Python (3.x) and necessary dependencies, plus an environment where you can run commands for both the MCP server and the web API.

Windows (Recommended)

Automated setup:

.\setup-windows.bat

This will create a virtual environment, install dependencies, generate a template .env file, and guide you through configuration.

Manual setup:

Create a virtual environment, activate it, install the package, and configure the environment file.

python -m venv venv
.\venv\Scripts\Activate.ps1
pip install -e .
cp .env.example .env
# Edit .env with your settings

Initialize the PostgreSQL database if you are using PostgreSQL, and start the servers as needed.

Quick commands to start components (Windows):

.\start-server.bat
.\start-mcp-server.bat
.\install-dependencies.bat
.\init-database.bat

Linux / macOS

Install dependencies and configure the environment, then run the MCP server and the web API.

pip install -e .
cp .env.example .env
# Edit .env with your settings

Run the components (Linux/macOS):

python -m cli.mcp_server
python -m web.server

Claude Desktop Configuration

Add the following MCP configuration to Claude Desktop so the MCP server can be launched from the desktop client.

{
  "mcpServers": {
    "fccs_agent": {
      "command": "python",
      "args": ["-m", "cli.mcp_server"],
      "env": {
        "FCCS_MOCK_MODE": "true"
      }
    }
  }
}

API Endpoints

The web API exposes endpoints to manage tools, executions, feedback, and metrics. Use these endpoints to discover available tools and run them as needed.

Available endpoints include health, tool listing, execution, per-tool invocation, feedback submission, and metrics retrieval.

Deployment

You can deploy using standard container or cloud options. The server supports Docker and cloud Run deployment patterns.

Docker example:

docker build -t fccs-agent .
docker run -p 8080:8080 --env-file .env fccs-agent

Google Cloud Run example:

gcloud run deploy fccs-agent \
  --source . \
  --region us-central1 \
  --allow-unauthenticated \
  --set-env-vars FCCS_MOCK_MODE=true

Feedback System

The agent tracks tool executions and supports user feedback. You can rate executions and review metrics via the API endpoints to improve performance over time.

Notes

The server provides mock mode and real FCCS integration options, along with bilingual support. Use the mock mode in development to simulate FCCS interactions without connecting to a live FCCS environment.

Available tools

get_application_info

Retrieve FCCS application details.

get_rest_api_version

Return the REST API version information.

list_jobs

List recent FCCS-related jobs.

get_job_status

Query the status of a specific job by ID.

run_business_rule

Execute a business rule within FCCS.

run_data_rule

Execute a data load rule within FCCS.

get_dimensions

List all FCCS dimensions.

get_members

Get members of a specific FCCS dimension.

get_dimension_hierarchy

Build and return a hierarchy for a dimension.

get_journals

List journals in FCCS.

get_journal_details

Fetch details for a specific journal.

perform_journal_action

Approve, reject, or post a journal.

update_journal_period

Update the period for a journal.

export_journals

Export journals data.

import_journals

Import journals data.

export_data_slice

Export a slice of grid data.

smart_retrieve

Perform smart data retrieval.

copy_data

Copy data between FCCS data slices.

clear_data

Clear data in a target area.

generate_report

Generate FCCS reports.

get_report_job_status

Check status of an asynchronous report.

export_consolidation_rulesets

Export consolidation rulesets.

import_consolidation_rulesets

Import consolidation rulesets.

validate_metadata

Validate metadata for FCCS objects.

generate_intercompany_matching_report

Create an intercompany matching report.

import_supplementation_data

Import supplementation data.

deploy_form_template

Deploy a form template.