OpenAI
ChatGPT
ChatGPT supports remote MCP apps in developer mode, including reading data and taking actions. We’ll check your account access and help scope a course workflow.
ChatGPT MCP documentation →Model Context Protocol · MCP
Bring ChatGPT and Claude
into your training workflow.
Read course data, create courses, update participants and send messages. Run modern AI workflows on CompetenceFlow’s ISO 27001-certified business backbone.
Your AI + CompetenceFlow
Your AI conversation
“Help me prepare next week’s courses.”
MCP
CompetenceFlowTraining recordsCourse context for your AI workflow
01Read course dataBring in the context
02Update participantsKeep records current
03Prepare messagesReview before sending
Review the details. Then take action.
The backbone for your AI workflows
A conversation that can
move the work forward.
Work with training records from a compatible AI application. ChatGPT and Claude offer conversational workflows; Microsoft Copilot Studio and Google Gemini CLI offer options for teams building their own agents.
ChatGPT, Claude & your AI toolsWhere you ask, plan and review
↔MCP
CompetenceFlowCourse records and training actions
Security for agentic workflows
Modern AI workflows.
A certified business backbone.
ISO 27001-certified information securitySee our security and certification details →
Gain the speed and flexibility of AI-driven workflows while keeping your training operation on CompetenceFlow’s ISO 27001-certified backbone. Your course records and core processes stay in a maintained business system.
A standalone vibe-coded implementation leaves your team responsible for building and maintaining the underlying data storage, permissions and security. CompetenceFlow provides that foundation, with EU hosting, role-based access, encryption and activity logs to protect your business data.
Use AI agents to plan courses, handle participant changes and prepare messages through MCP. Your team can improve the workflow without rebuilding the system that runs the business.
During setup, we agree which data your chosen AI tool can access and which actions need your team’s review.
Choose your AI tool
ChatGPT, Claude,
Copilot Studio and Gemini CLI.
These applications support MCP. We’ll confirm the connection, permissions and available actions for your account during setup.
Anthropic
Claude
Claude supports custom connectors to remote MCP servers. Bring course information into a conversation, review the next step and use the available training actions.
Claude connector documentation →Microsoft
Microsoft Copilot Studio
Copilot Studio lets teams add MCP tools to custom agents. We’ll review the connection and your organisation’s permissions for a training workflow.
Copilot Studio MCP documentation →Google Gemini CLI
Gemini CLI supports MCP servers for teams working from the command line. It is an option for technical teams building their own course workflows.
Gemini CLI MCP documentation →Available through MCP
Four ways to work with CompetenceFlow.
Read
Course data
Bring course information into your AI conversation for planning and follow-up.
Create
Courses
Use your AI tool to create course records in CompetenceFlow.
Update
Participants
Make participant changes through the connection, without re-entering them in another tool.
Send
Messages
Prepare communication in your AI workflow and send it through CompetenceFlow.
Put it to work
Start with a job
your team does every week.
These examples show how the available MCP actions can fit into a working day. Bring your own course process to the demo and we’ll walk through it.
Set up the next course.
Read existing course information, work through the new dates and details in your AI tool, then create the course in CompetenceFlow.
Example request: “Help me prepare the next run of our first-aid course.”
Deal with the changed attendee list.
Use the course context to review the required changes, then update participant records through MCP.
Example request: “Help me update the participants for this company booking.”
Prepare and send the instructions.
Use the course details to prepare a message. Check its recipients and wording, then send it through CompetenceFlow.
Example request: “Prepare joining instructions for next week’s course for me to review.”
Getting connected
Bring the tool.
We’ll bring the course workflow.
Tell us which AI application you use and what you want it to do. We’ll show the connection and available actions, then help you set up a first workflow.
Setup guidance is available from our team. Public documentation is not yet available.
For background, read the MCP introduction and the application guides above. Include your chosen AI provider in the security and data handling review.
Questions about MCP and AI workflows
What is MCP?
MCP stands for Model Context Protocol. It is a standard that lets AI applications connect to external data and tools. CompetenceFlow uses it to make training data and actions available to compatible AI assistants.
Can I use CompetenceFlow with Claude?
Yes. CompetenceFlow supports MCP for use with Claude and other compatible AI tools. Available actions include reading course data, creating courses, updating participants and sending messages.
Is this a separate training database?
No. MCP is a way for your AI tools to work with CompetenceFlow. Course and participant records remain part of the same training operation your team manages in the application.
Why use CompetenceFlow instead of a standalone vibe-coded training system?
A vibe-coded prototype can help you test an idea. Making it your business system also means taking responsibility for data storage, permissions, security updates and ongoing maintenance. CompetenceFlow provides an ISO 27001-certified backbone for your training records and core operations, with EU hosting, role-based access, encryption and activity logs. Build AI workflows around that foundation and focus on the work you want to automate. Your chosen AI tool’s access and data handling are agreed during setup.
Where can we find setup instructions?
Contact our team for setup guidance. There is no public setup guide yet. We’ll walk through the connection, available actions and a workflow relevant to your business.
Can I use ChatGPT for training management through MCP?
ChatGPT supports remote MCP apps in developer mode, with both read and write tools. CompetenceFlow provides course data and actions through MCP. We’ll check your ChatGPT account, authentication and available actions before setting up the workflow.
Can Microsoft Copilot connect to CompetenceFlow?
Microsoft Copilot Studio supports adding MCP servers to custom agents using Streamable HTTP. We can review that route for your training workflow, including authentication and organisation policies. It requires a configured Copilot Studio agent.
Can I use Google Gemini with MCP?
Google’s Gemini CLI supports MCP connections for command-line workflows. We can review that option with your technical team. The setup described here is for Gemini CLI; connection options depend on the Google product you use.
Show us what you want your AI to handle.
We’ll demonstrate how CompetenceFlow’s course data and actions can support that workflow.