Using MCP Server for Technical Writer’s Productivity

Anthropic developed the Model Context Protocol (MCP) to enable AI agents to interact with one another. The MCP architecture involves an MCP client that discovers services offered by many MCP servers to which it is connected. The MCP client discovers services offered by other MCP servers by establishing a connection via HTTP request and understanding how to utilize different tools. The MCP client orchestrates calls to different tools hosted by different MCP servers. The MCP client and MCP server use a Large Language Model (LLM) under the hood to orchestrate tool calls. The tools are API endpoints with detailed descriptions about their purpose and parameters. Even though the MCP looks technical, it enables intelligent workflow automation. It has the power to bypass the entire UI layer of your business application, and it can accomplish tasks through the MCP client.

How Is It Relevant to Technical Writers?

Technical writers undertake many activities as part of the DDLC (Document Development Life Cycle) frameworks. This includes research, collecting information from various stakeholders, writing content, adding article metadata, publishing, and so on. Once the article is published, technical writers also undertake various activities to collect feedback from customer-facing teams and look at customer support tickets to identify knowledge gaps. Most of these activities are time-consuming and require the collation of information from different business applications. Technical writers do not have access to many business applications, thus relying on stakeholders to provide inputs and metrics. With the advent of the MCP server, this can be handled more effectively. The technical writers can connect to various business systems using their MCP server with their client application.

Use Case 1: Research and Writing

To undertake research, technical writers need access to internal documents stored in many knowledge base systems such as SharePoint, OneDrive, and so on. In addition to this, they also need access to emails and other artefacts stored in online storage systems. With a click of a button, technical writers can configure the MCP servers of respective business systems in their MCP client, such as Claude Desktop.

List of MCP connectors available in the Claude Desktop client

Many LLM vendors have a deep research mode, which can augment internal articles with external knowledge sources. This would help technical writers to do comprehensive research and cover topics thoroughly. More importantly, new knowledge can be written to any knowledge sources directly from the MCP client via respective connectors.

Use case 2: Workflow Automation

Technical writers interact with product managers for every product feature release. Sometimes product features may not ship on the scheduled time, and technical writers are informed at the eleventh hour to fix the release notes. These tasks are urgent and important. With the MCP client, technical writers can automate workflow with a simple prompt. They can add, modify, and remove content from their release notes article quickly with a prompt.

Workflow showing content updation in the release note

Technical writers can now connect to the customer support ticketing system via MCP connectors regarding any particular issue and compare it with the existing knowledge base to identify and fix content gaps. If the customer support system has some inherent knowledge that is not available in the customer-facing knowledge base, technical writers can quickly identify these scenarios and update the relevant content more quickly. The MCP client can intelligently orchestrate workflow between the knowledge base and customer support applications.

Streamline research, content updates, and workflow automation with an intelligent knowledge base. Give Document360 a try!

GET STARTED

Why Is Human-in-the-Loop Important?

The MCP client can search, add, update, and delete content based on certain insights. It is very important to have human-in-the-loop processes in place, not to slow down the process but to enhance trust. Technical writers will have to validate the information based on different sources, eliminate ambiguity in the content, and ensure that new information serves as the truth. Technical writers will elevate to an “AI manager” who oversees

• AI agents working and their performance
• Evaluate the responses generated by the GenAI chatbot
• Evaluate the effectiveness of task completion by the MCP client/AI agents to attribute the value of documentation
• Review and validate the AI-generated content from different sources.

The MCP client and server will empower technical writers to access a multitude of business applications and enable information flows more effectively. It will harness the power of tacit knowledge and add it to the knowledge repository instantly. With the MCP, it becomes easier for technical writers to share new information with stakeholders and accomplish tasks. The table below shows the changing characteristics of technical writers’ roles after the introduction of MCP.

This new paradigm also enables technical writers to accomplish tasks and do their work in the MCP client rather than logging into separate business applications. The technical writers will review the AI-generated content. Technical writers will primarily review AI-generated content and make it easy for AI agents to utilize content facilitated by the MCP server.

The Takeaway: How MCP Server Transforms Documentation Workflows

The MCP server is still in the nascent stage, but many software providers are adding support for MCP inside their products. The LLM vendors, such as Anthropic and OpenAI, have added support for MCP and are promoting a marketplace for MCP servers that can be added to their MCP client. The MCP is a paradigm shift in terms of how value is created by technical writers and how this is captured in terms of business outcomes.

The MCP approach automates and intelligently orchestrates many of the processes of technical writing practices. It will reduce the stakeholder interaction with many business system owners, thus empowering technical writers to access their business systems for intelligent work automation.