HappySupport
MCP and WebMCP

Your documentation as an MCP server

AI agents shouldn't have to scrape your help center. HappySupport exposes your docs as callable tools, and lets your team write and maintain articles through MCP too.

These teams document with HappySupport

FlipjupusPeakboardaworkneuroflash

Agents already read your documentation. The problem: you don't get to decide what they read or how they understand it. They take the page exactly as they find it.

The fix isn't another export. Agents need real tools where you decide how they search and read. And your team plugs HappySupport into its existing workflow through MCP.

WebMCP

Agents query your docs instead of scraping them

Semantic search results in the help center

HappySupport registers WebMCP tools on your help center. A WebMCP-capable agent in the browser can search your articles by meaning, read one in full, list everything published, and pull up related articles.

So the agent stops guessing and gets exactly the ranking you curate for your readers: the right article, the full text, and a link it can cite as the source. It costs you no extra upkeep, and browsers without WebMCP still see a perfectly normal help center.

MCP for your docs team

Write and maintain articles from your own agent

A new article being drafted in the HappySupport editor

The people maintaining the documentation get MCP too. Point Claude, Cursor or your own agent at HappySupport and it finds the right article, drafts new ones and keeps existing ones current, without leaving the tool you work in.

Every change lands in the same editor, with the same versioning and review your team already uses. The agent proposes, a human accepts.

What agents can call

Four tools that show agents exactly what they need

Every tool is described for the agent, so it picks the right one without guessing. Plus access control and the write path for your team.

Semantic search

Search by meaning and keyword, ranked the way your help center ranks.

Full article

Fetch the complete markdown of one article by its id.

Related articles

Follow one article to the ones next to it.

Create and update

Your docs team drafts and maintains articles through MCP.

Access rules apply

Gated articles stay gated. An agent sees what the reader sees.

Read-only by default

Article text reaches the agent as data, never as instructions.

Related

Docs an agent can actually use

MCP is the interface. This is what sits behind it.

Built for

Teams shipping agents

Your help center, ready before your coffee.

Start for free and publish your first article today.

HappySupport