The Creative Force MCP Server connects your AI assistant directly to your studio's live production data. Ask a plain-language question and get a live answer– no dashboard navigation, no exports, and no code. This article explains what the MCP Server is, what you can do with it, and the AI tools it supports.
🧪 Open Beta: The Creative Force MCP Server is currently in Open Beta. Supported AI tools and available features may change. To share feedback or report an issue, ask your AI assistant to use the Send Feedback tool.
MCP Overview
Modern AI Assistances (Claude, ChatGPT, Copilot Studio) are highly capable, but only know what they can access and communicate with. Without a connection to Creative Force, getting data out of the platform means logging in, navigating to the right screen, and often exporting it manually.
Creative Force's MCP Server changes that. It acts as a secure bridge between your AI assistant and Creative Force, translating plain-language questions into live data queries and returning answers directly in your chat– using the same data and the same permissions as your Creative Force account.
Supported AI Tools
AI Tool | Interface | Availability Status |
Claude by Anthropic | Desktop App Web App Code Extension | Available Available Available |
ChatGPT by OpenAI | Web App | Available |
Copilot by Microsoft | Web App | Coming Soon |
Getting Started
The MCP Server documentation is split into three distinct knowledge pillars, and they should be read and acted on in a specific order. Each article in the Collection is outlined below in the order in which they should be reviewed.
Knowledge Pillar | Information |
MCP Server: Overview
You're here. | The Overview article covers the general scope of what an MCP Server does, how that relates to Creative Force, and introduces other relevant resources. |
MCP Server: Queries
| The Queries article covers the necessary permissions for MCP access in Creative Force, as well as tips for querying once setup is completed. |
MCP Server: Setup
| The setup article covers the necessary setup steps outside of Creative Force, broken down by each supported AI Tool. |
The overall process of learning about, setting up, and participating in the MCP Server Open Beta can be broken down into a few distinct steps, outlined below.
Step 1 → | Step 2 → | Step 3 → | Step 4 → | Step 5 |
Read the Overview article to understand MCP scope. | Enable MCP permissions to users via the Studio Settings. | Connect your preferred AI Tool to the Creative Force MCP Server. | Use your AI Tool of choice to query your production data. | Submit feedback to MCP team to contribute to future updates. |
You're here. |
Glossary
Term | Definition |
MCP (Model Context Protocol) | An open standard that allows AI assistants to connect directly to external software and retrieve live data on your behalf. |
MCP Server | The component that sits between your AI assistant and Creative Force. It receives requests from the AI, calls the relevant API, and returns results in a format the AI can understand. |
MCP Client | The AI tool that connects to the MCP Server– for example: Claude, ChatGPT, or Copilot Studio. |
OAuth | A secure, industry-standard login protocol. When you connect Creative Force to an AI tool, OAuth handles authentication in the background. You log in once, and your access token is cached for future sessions. |
Access token | A secure, temporary credential stored on your device after login. It proves to the MCP Server that you are authenticated without requiring your password on every request. Tokens expire periodically and are refreshed automatically by your AI tool. |
Available Functions
Data Queries
Once connected, the Creative Force MCP Server gives your AI assistant access to query data covering your live production data. You can ask your AI assistant natural language questions about your production data. Available capabilities include:
Category | Functionality |
Production Data | Query asset counts, production status, and photographer activity |
Jobs & Deadlines | Track job status, find overdue work, and check what's due this week |
Tasks & Workflows | Monitor task progress by step, assignee, or status across your production workflow |
Sample Tracking | Find where samples are, which haven't been returned, and which are overdue |
Planning & Scheduling | Check upcoming shoots, session alerts, and whether your team is on track |
Product Catalogue | Look up individual products, filter by category, and check production status |
Limitations
Read-Only: The MCP Server is currently read-only. Your AI assistant can query and report on your data, but cannot create, edit, or delete any records. Write access is planned for a future release.
BI Tools: MCP connects to AI assistants only. It does not integrate directly with BI tools such as Power BI or Looker.
Interface: MCP requires a connected AI tool and is not a standalone interface.
Planned Expansions
Creative Force is hard at work on future implementations and looks forward to sharing further functionality as it becomes available. In the table below, you can review some currently planned implementations to look forward to.
Planned Feature | Overview |
Query: Studio Settings | Query your settings to better understand how your instance's configuration works across your data. |
Query: Workflows | Query data in your Workflows and Styleguides to better understand your production processes. |
Write: Planning | Use your AI Tool to manage your Planning data– including creating and editing existing Sessions. |
MCP Queries
Crafting Prompts
AI Agents connected to an MCP Server have the ability to check and source a large amount of data. But how it checks for that data is also dependent on how you communicate your needs. In the sections below, we will cover a few general tips for getting the most accurate responses from your MCP-Connected agent.
Define Details
To get the most accurate and useful responses, it helps to be specific and outcome-focused in how you phrase your prompts. Rather than asking broad questions, try to frame your prompt around a clear, concrete need:
"How are my products doing?" ❌
"Which products produced in the last 3 months have fewer than two approved images?" ✅
"Which Samples still need model shots?" ✅
The more precisely you describe what you're looking for, the better the agent can identify the right data to surface.
Full-Scope Prompts
When your question involves multiple variables, such as a specific Workspace, Job, or deadline, include all relevant identifiers in a single prompt. For example, instead of asking two separate questions, combine them:
"Which products assigned to Workspace X are missing their final delivery assets and due within the next 7 days?" ✅
This gives the agent a full picture of what you need and reduces back-and-forth. If you're unsure of exact product names or IDs, describe the context as clearly as you can (e.g., by style type, Category, or Workflow step), and the agent will do its best to match your intent against available production data.
Context VS Limitation
If a response seems incomplete or unexpected, try rephrasing with a bit more context or breaking a complex question into smaller, sequential asks. When a limitation is encountered, the AI agent should be able to clearly state what that limitation is– so you know that it is not a context issue.
Sending Feedback
While using the Creative Force MCP Server, you can ask your AI Tool to send feedback directly to our team. For example, when encountering a limitation, you can use a prompt like below to loop us in and include the relevant information automatically:
Send this to Creative Force as feedback. Add my prompt and use case for them to understand.
When submitting feedback, please be sure to include the business use case and the specific limitation you are facing.
Next Steps
Now that you understand the scope of how the Creative Force MCP server works, you can move forward to reviewing and enabling permissions for MCP access in Creative Force.