How MCP works
Essentially, MCP enables an AI agent to access specific tools and actions. The agent knows its possible functions, recognizes which action to perform based on the user's request, and then translates the result into a response that humans can understand.
Controlled automation for AI agents.
Model Context Protocol (MCP) describes a standardized approach that enables AI agents to perform deterministic actions in a secure and controlled environment. While AI models are inherently non-deterministic, MCP creates a clear interface between intelligent decision-making and deterministic automation.
Action under control
AI agents can only trigger previously approved and defined workflows in a targeted manner. Although the execution of these automation workflows allows operational intervention in the infrastructure, this is only possible within a clearly defined, controlled framework.
Automation meets language
The automation platform acts as a deterministic and controlling execution layer, while MCP forms the bridge between natural language and technical automation.
Seamless integration into your AI landscape.
The DICOS MCP Server seamlessly integrates the APIs of the AWX Automation Platform into your AI landscape and is fully configurable. The DICOS MCP Server acts as an intermediary and provides the AI agents with the desired automated tools.
This integration layer can be used to provide both internal platform functions and complete automation workflows themselves.
For the user, this means that automation services are available directly via a natural language interface. When a user interacts with an AI chatbot, the AI agent can decide which MCP tool to call up based on the request.
Practical example: AWX
Via chat to the DNS entry.
AI helps you solve complex infrastructure tasks. We show how easy this can be in the following example.
1
Assignment via chat prompt
A user asks the AI agent to create a new DNS entry for a specific IP address.
2
Execution in AWX
The AI agent calls up the corresponding MCP tool, which executes the predefined automation with the AWX automation platform.
3
Reply via chat
The AI agent communicates the progress and results to the user in an easy-to-understand manner via chat.
This creates a clear separation between dialogue, decision logic, and execution—secure, traceable, and fully automated.
Automation is no longer perceived as a technical hurdle, but rather as an intuitive service that offers direct added value.
Dr. Tobias Bautze, Solution Architect at DICOS
Multi-agent architecture with clear tool assignment
The DICOS MCP Server enables multi-agent operation with fine-grained task separation between different AI agents. This allows multiple AI agents with their defined actions to access different tools provided by the MCP server. The agents each act within their assigned area, which makes all actions traceable and provides the prerequisite for a modern, scalable architecture.
Automation with AI
Advantages of an AI agent
The use of an AI agent as an interface to the AWX Automation Platform offers numerous advantages for both IT teams and specialist departments outside of IT.
Acceleration and simplified operation
AI agents enable access to complex automations via natural language. Tasks that previously required multiple steps or deep expertise can now be triggered in seconds.
Reduction of sources of error
The AI agent abstracts technical details such as playbooks, job templates, or API calls. Users interact on a professional level, while the technical implementation takes place in a controlled manner in the background. This reduces the error rate.
Transparency and traceability
MCP can be used to query status information, job results, and reports in a targeted manner and present them in a clear format. Users can keep track of the progress and results of automation processes.
Automated processes with Ansible and AWX
Automation doesn’t have to be complicated. We help you take the first steps toward automating your IT processes with AWX and Ansible.
Safe and risk-free use of AI
No compromises on governance and compliance
When companies integrate artificial intelligence into their IT automation, they must place particular emphasis on data protection, compliance, and security. Especially in productive use, sensitive information must remain protected at all times and regulatory requirements must be met.
A key foundation is the use of AI stacks that can be operated entirely on-premises, because only in a local, controlled environment can sensitive data be reliably protected, compliance requirements met, and internal security policies consistently enforced.
Another crucial aspect is the clear definition of guidelines for the use of AI agents. Companies must specify exactly which actions the AI is allowed to trigger and under what conditions. Access to MCP tools and automations can be controlled granularly and linked to existing standard processes.
This ensures that AI agents are only allowed to perform predefined, tested, and approved actions. This enables the secure use of AI in productive scenarios within the specified framework of governance, compliance, and security.
More about AI
Automation x AI
MCP Server for your AWX Automation Platform
Want to get more out of your AWX Automation Platform?
Get in touch with us. We will advise you.
Get in touch with us. We will advise you.