AI Coding Assistants

GitHub Copilot, Cursor, Claude Code and similar assistants write code from what they know and what they can look up. For TX Text Control, Text Control gives them two sources: the Text Control MCP Server answers their questions from the current documentation, blog and product pages, and Agent Skills give them step-by-step workflows for common tasks. With both in place, a request such as "create a Document Editor application" doesn't need a long prompt that explains the API first.

Both tools help you write your application; they don't run inside it. Adding language models to the application itself is the subject of Agent Workflows and the other pages of AI Integration.

Two Tools for Coding Assistants

The Text Control MCP Server is a different product from the MCP Document Server, which you host yourself and which creates and edits documents.

The Text Control MCP Server

The Text Control MCP Server is a public, read-only endpoint at https://www.textcontrol.com/mcp/v1. It needs no authentication, returns only content that is already public on the Text Control website and can't change any data. It uses the same data as the website, so new blog articles or documentation changes are available immediately.

The server implements the MCP specification of 2025-03-26 with the Streamable HTTP transport: a client posts JSON-RPC 2.0 requests to the endpoint and receives JSON. Four methods cover the whole exchange:

The tools are grouped into eight domains, each with search, list and get operations:

Search and list tools return up to 15 results by default and at most 25. Ask for a number of results to set their limit; larger result sets are paged with offset. When you ask for the full content, the get tools include the complete body as Markdown.

Connect Your Assistant

Add the endpoint to the MCP configuration of your assistant. In Visual Studio Code with GitHub Copilot, add it to .vscode/mcp.json in your workspace, or to the user configuration that the MCP: Open User Configuration command opens:

{
  "servers": {
    "textcontrol": {
      "type": "http",
      "url": "https://www.textcontrol.com/mcp/v1"
    }
  }
}

Claude Code reads the same entry from .mcp.json in your project, or from ~/.claude.json for all projects, under mcpServers:

{
  "mcpServers": {
    "textcontrol": {
      "type": "http",
      "url": "https://www.textcontrol.com/mcp/v1"
    }
  }
}

Other assistants use their own files and keys:

For other scopes and further settings, see the documentation of your assistant.

Ask About TX Text Control

Once connected, ask in natural language; the assistant chooses the tools:

The documentation tools filter by product with product_name_id. For this product, the ID is tx-text-control-dotnet-server.

Agent Skills

Agent Skills are an emerging open standard for extending AI agents with reusable workflows and domain knowledge. A skill is a folder with a SKILL.md file and, usually, a references folder:

• SKILL.md starts with front matter that names the skill and describes when to use it. The body lists what the skill can do, example requests with the matching reference file, the workflow the assistant follows and its rules.
• references holds one file per task, with the goal, the source article, the prerequisites, the code and notes.

The front matter of the mail merge skill starts like this:

---
name: txtextcontrol-servertextcontrol-mailmerge
description: |
  Use this skill when a user asks how to build ServerTextControl-based
  MailMerge workflows with TX Text Control .NET Server.

  Trigger it for requests involving JSON data merges, XML data merges,
  object and list merges, merge blocks, skipping rows during merge,
  and exporting merged output to PDF.
metadata:
  author: Text Control

The assistant scans the skills in your workspace, loads the one whose description matches a request and follows its workflow. The rules of the Text Control skills tell it to use only the TX Text Control APIs of the reference files, not to invent APIs, credentials or license keys and not to hardcode license keys.

The Text Control Skills

For the Document Editor, start with the setup skill of your platform and then use the API skill. The reference files of the ServerTextControl skills are based on validated code from the Text Control blog and documentation.

Install the Skills

1. Clone the repository, or download it as a ZIP file:

   git clone https://github.com/TextControl/txtextcontrol-servertextcontrol-skills
   

2. Copy the skill folders you need into a folder that your assistant scans: .github/skills, .claude/skills or .codestudio/skills in your workspace:

   your-workspace/
   ├── .github/skills/
   │   ├── txtextcontrol-servertextcontrol-mailmerge/
   │   │   ├── SKILL.md
   │   │   └── references/
   │   └── txtextcontrol-servertextcontrol-tables/
   │       ├── SKILL.md
   │       └── references/
   └── your-project-files...
   

3. Start a new chat session, so that the assistant discovers the skills.

Use a Skill

Describe the task, for example "Create a simple ServerTextControl MailMerge application using a DOCX template and JSON data." The assistant loads the matching skill by itself. When your assistant supports slash commands for skills, you can also name the skill:

/txtextcontrol-servertextcontrol-mailmerge create a JSON template merge flow

Use Both Together

The MCP server and the skills complement each other. Asked to create a Document Editor application, an assistant with both loads the setup skill for the steps and queries the MCP server for current API documentation and samples to fill in the details. The blog article AI Tools: MCP Server and Agent Skills describes this combination.

Review the generated code as you would any other code, and build and test it against the version of TX Text Control you use.

Next Steps

• Document Editor with ASP.NET Core: set up the Document Editor step by step yourself.
• MCP Document Server: host document tools for AI clients in your own application.
• Examples: the skill repositories and the other AI samples.