Add MCP (Model Context Protocol) server support for Otter Blocks

Impact: 60 · Confidence: 95 · Complexity: 85

Description

Problem

Users utilizing AI assistants (like Claude) cannot effectively automate the creation of WordPress pages using Otter Blocks because the AI lacks context about Otter's specific block structures, schemas, and attributes. While Neve has an MCP integration for theme-level AI building, the granular block-level generation from HTML/CSS prototypes to Otter Blocks requires an Otter-specific MCP server.

Desired Behavior

Provide an MCP (Model Context Protocol) server/skill for Otter Blocks that exposes the plugin's block capabilities, schemas, and attributes to compatible AI assistants (e.g., Claude, generative AI IDEs like OpenCode). This will enable AI to accurately write and convert HTML/CSS prototypes into valid Otter Blocks WordPress markup.

Acceptance Criteria

• Develop an MCP server for Otter Blocks that defines block schemas and creation parameters.
• Output valid block markup that successfully renders in the WordPress block editor when generated by an AI assistant.
• Provide documentation on how users can connect their AI assistants to the Otter Blocks MCP server.

Customer Context

The customer is a tech-savvy Neve Pro Agency user who uses Claude and OpenCode. They want to automate their workflow by converting HTML/CSS prototypes into WordPress pages built with Neve Pro and Otter Blocks Pro. They are specifically asking for an AI skill (MCP) for Otter Pro to make this possible. Additionally, they reported that some starter site links on the website are currently broken.

Root Cause Analysis

The root cause is a gap in AI integration capability for Otter Blocks. To correctly programmatically generate WordPress blocks, AI assistants need strict schemas and context about block attributes, otherwise, they generate invalid or basic HTML. The customer wants to leverage AI (specifically through the Model Context Protocol, which Neve just adopted) to generate Otter Blocks layouts from standard HTML/CSS, but Otter currently lacks the MCP server required to supply the AI with the necessary block structure definitions.

Reasoning

The customer is asking for an AI 'skill' (MCP server) for Otter Pro to allow AI assistants like Claude to generate valid Otter Blocks markup from HTML/CSS prototypes. While Neve recently released an MCP integration, no such integration exists for Otter Blocks according to documentation and GitHub searches. This is a net-new feature request for the Otter Blocks repository. The customer also noted broken links on the starter sites page, which I've captured in the context.

Source: HelpScout #3372213285
Generated by feature-request-triage workflow (ID: feature-request-triage_6a43f6258e72d1.07444311)

No implementation files, tests, or entry points are identified. Start by reviewing the existing Neve MCP integration referenced in the issue and the Otter Blocks schemas and attributes; done means an MCP server can generate valid Otter Blocks markup, with connection documentation for compatible AI assistants.