Building WordPress Tables via AI Chat: A FlexTable MCP Hands-On

FlexTable's free plugin uses MCP to let AI assistants like Claude and ChatGPT build WordPress tables in natural language.
FlexTable is a free WordPress table plugin with a built-in MCP server that lets AI assistants like Claude, ChatGPT, and Cursor interact directly with a WordPress site. Users enable MCP, paste a setup prompt, and describe their needs in natural language. The AI reads Google Sheet data, builds table structures, applies styles, and places tables on new pages—no manual work required. Generated tables are saved as drafts for human review before publishing, lowering the risk of errors. The MCP protocol layer also leaves room for future plugins to enable broader AI-driven workflows.
If you build tables in WordPress regularly, you know the drag-and-drop, style-tweaking, preview-and-publish grind. A free plugin called FlexTable aims to change that: through the MCP protocol, you can have Claude, ChatGPT, or Cursor build tables for you.

What Is FlexTable MCP?
FlexTable is a free WordPress table plugin whose core innovation is a built-in MCP (Model Context Protocol) server. MCP is an open protocol that lets AI assistants communicate with external tools, and FlexTable uses it to expose table-building capabilities to AI.
That means you don't have to click around the WordPress admin manually—you just tell the AI what you want in natural language. The plugin itself is free, lowering the barrier to trying it out.
MCP (Model Context Protocol) is a standardized protocol proposed and open-sourced by Anthropic in late 2024. It aims to let AI assistants (such as Claude and ChatGPT) call external tools and data sources in a unified way. It uses a client-server architecture: the AI application acts as the client, external services (such as WordPress plugins, databases, and APIs) act as servers, and the two communicate via JSON-RPC messages. The core value of MCP is "implement once, reuse everywhere"—tool developers only need to build a single MCP server, and any MCP-compatible AI client can call it, with no need to develop a separate integration for each AI platform. This differs from traditional Function Calling, which is typically tied to a specific model or platform. Today a large number of open-source MCP servers cover scenarios such as the file system, Git, and databases, and FlexTable is an attempt to bring this model into the WordPress ecosystem.
Three Steps to Get Started: From Enabling MCP to Generating Tables
The whole flow is designed to be quite simple, and the official docs boil it down to a few actions.
Step 1: Enable the MCP Server
Turn on the MCP server switch in the FlexTable plugin settings. This is the necessary step that lets AI assistants access the table-building capabilities.

Step 2: Paste the Setup Prompt into Your AI Assistant
Once MCP is enabled, the plugin provides a setup prompt. Copy this prompt and paste it into the AI assistant you use to complete the connection configuration. Supported assistants include Claude, ChatGPT, and Cursor.
The setup prompt here plays the role of a "connection handshake." Because AI assistants such as Claude and ChatGPT don't know by default that the FlexTable MCP server exists, the plugin generates a piece of text containing the server address, authentication information, and a list of available tools. By pasting it into the AI assistant's conversation, the user is effectively telling the AI, "Here's an available table-building tool, and here's how to call it." For clients with native MCP integration (such as Cursor), this prompt may directly trigger a connection at the protocol layer; for ordinary chat interfaces, it relies more on the instructions in the prompt to help the AI understand how to construct requests. This also explains why the experience may differ slightly across assistants—clients with native MCP support can achieve more automated and more reliable tool calls.
Step 3: Just State What You Need
Once configuration is complete, you only need to describe the table you want in natural language. For example, you can ask the AI to read data from a Google Sheet and then build a table based on that data.

Real Capabilities: How Far Can the AI Go?
According to the official demo, after receiving instructions Claude can carry out a series of connected operations: read a Google Sheet, build the table structure, apply styles, and place the table on a new page.

The entire process requires no manual intervention from you. The AI doesn't just generate the table's HTML code—it interacts directly with the WordPress site, completing the full chain from data source to page presentation.
Safety Mechanism: Drafts and Publishing Are Separate
The biggest worry with automation is that a misstep will push content live directly. FlexTable has a clear design here: tables generated by the AI land as drafts, so you can review the result first and undo it if you're not satisfied.
Nothing is visible externally until you click the publish button. This "draft-first" mechanism turns AI table-building from a "show-off trick" into a production tool you can use with confidence.
In WordPress, a draft is an unpublished content state that is visible only to logged-in users with editing permissions, and does not appear on front-end pages or in search results. By having AI-generated tables land as drafts, FlexTable effectively combines WordPress's native content state machine with AI operations: the AI has permission to create and edit drafts, but publishing permission remains in human hands. This "human-in-the-loop" design is a common safety pattern for automation tools—it retains AI's efficiency while avoiding irreversible public errors. In addition, the fact that drafts can be undone means that even if the AI generates a table with structural errors or messy styling, users can easily roll it back before publishing rather than facing the pressure of fixing content that's already live.
What It Means for WordPress Workflows
The WordPress ecosystem has long lacked an elegant table solution. Traditional approaches rely either on page builders' table modules, on hand-writing HTML, or on bloated table plugins. By using MCP to hand table-building over to AI, FlexTable is fundamentally changing how content production is interacted with.
One detail worth noting: MCP's existence as a protocol layer means this model has room to expand. If more WordPress plugins follow suit and support MCP, a workflow where you "build an entire site's content with one sentence" could emerge in the future. Of course, that depends on how fast the ecosystem adopts it and on AI's ability to understand complex sites.
Who Should Try It?
Content operators who frequently need to publish data tables, WordPress developers who want to reduce repetitive work, and AI tool users already using Claude or Cursor are all worth giving it a try. With a free plugin and a draft-protection mechanism, the cost of experimentation is low.
Currently the solution relies on Google Sheets as one of its data sources; users with other data sources should keep an eye on future feature updates.
One-line Summary
FlexTable, a free plugin, uses the MCP protocol to let AI assistants like Claude and ChatGPT build WordPress tables directly in natural language.
Paragraph Summary
FlexTable is a free WordPress table plugin whose core innovation is a built-in MCP server that enables AI assistants such as Claude, ChatGPT, and Cursor to interact directly with a WordPress site. Users only need three steps: enable the MCP server, paste the setup prompt into the AI assistant, and describe their needs in natural language. The AI can read Google Sheet data, build the table structure, apply styles, and place it on a new page—all without manual operations. On the safety side, generated tables are saved as drafts and require human confirmation before publishing, reducing the risk of missteps. The solution changes how WordPress content production is interacted with, and the MCP protocol layer leaves room for more plugins to follow suit and enable broader AI-driven workflows in the future.
Related articles

Anthropic's IPO Eve: Five Risks Behind the Trillion-Dollar Valuation and the Truth About Its Governance
Breaking down Anthropic's IPO prospectus: $184M daily revenue at a trillion-dollar valuation, the truth behind $104B in paper losses, a five-class share governance structure, unusual disclosures in 94 pages of risk factors, and how the lock-up schedule may affect retail investors.

FBI Warns: Hackers Weaponize AI Agent ARTEX to Attack 200+ Organizations
FBI warns that AI penetration framework ARTEX is being used to attack 200+ organizations, with CrowdStrike tracking the VaporPanda group. We break down ARTEX's agent orchestration architecture, the real attack chain, loss data, and the FBI's five defense recommendations.

FBI Warns: Free AI Hacking Tool Artex Spirals Out of Control, Breaches Cloud Systems in Three Hours
FBI warns that the free open-source AI penetration tool Artex is being abused by criminals and state-backed hackers to breach cloud systems in as little as three hours. This article analyzes real attack cases, overlooked side-door risks, and defense measures enterprises and individuals can take immediately.