Skip to content
ADVPL MCP
SkillsLogin

ADVPL context for your AI assistant

Totvs ADVPL MCP connects AI assistants to the knowledge and tools an ADVPL developer uses every day.

Create account
Learn more

What it is

  • An MCP server

    Implements the Model Context Protocol, the open standard that connects AI assistants to external context sources.

  • ADVPL context

    Gathers ADVPL documentation, conventions and examples so the assistant answers with the language's own vocabulary and practices.

  • Exposed tools

    Reads documentation on the ADVPL/TLPP language and framework, Protheus modules, reports, TIR tests, tax legislation, TSS and TCloud, and queries the Protheus data dictionary.

Why use it

  • More accurate answers

    The assistant stops guessing ADVPL syntax and conventions and queries real language context instead.

  • Less context switching

    Stay in your editor or chat with the assistant instead of constantly jumping to documentation.

  • Adopt without rewriting

    The MCP connects to the assistant you already use; no existing ADVPL code has to change.

When to use it

  • While writing ADVPL

    Ask the assistant for ADVPL syntax, functions and patterns as you code.

  • While reading legacy code

    Use the assistant to explain existing ADVPL routines with the right language context.

  • While onboarding a new project

    Set up the MCP at the start of a project so the whole team shares the same ADVPL context.

Where it works

  • In your AI assistant

    The MCP runs as an independent server that any protocol-compatible client can reach.

  • Supported clients

    Any agentic or coding tool that supports MCP servers, such as VS Code, Claude, Codex and Kiro.

  • No change to your environment

    You keep the editor, terminal and ADVPL workflow you already use.

How to set it up

  1. Install a compatible MCP client.
  2. Open the client's configuration file.
  3. Add the Totvs ADVPL MCP with the snippet below.
  4. Restart the client and confirm the server shows as connected.

MCP client configuration

"mcpServers": {
  "totvs-advpl-mcp": {
    "url": "https://totvs-advpl-mcp.engpro.totvs.io/mcp",
    "headers": {
      "Authorization": "SEU_TOKEN"
    }
  }
}