Documentation MCP

We provide an MCP (Model Context Protocol) server so your AI agent can search and read these docs interactively instead of relying on static context.

This MCP server exposes documentation only — it does not call the Boring Marketing API. Use it to let your agent search and read these docs. To actually run pipelines, fetch opportunities, or generate content, your agent must call the REST API directly with an API key. There is no API MCP server today.

Setup

Add the docs as an MCP server:

claude mcp add --transport http boring-marketing-docs https://docs.boringmarketing.com/_mcp

Now when you ask Claude about the Boring Marketing API, it can search and read the documentation directly.

Available tools

The MCP server exposes two tools:

searchDocs

Search documentation by keyword. Returns ranked results.

Parameter	Type	Required	Description
`query`	string	Yes	Search terms (e.g., `"polling runs"`, `"authentication"`)
`limit`	number	No	Max results, default 10, max 50
`type`	string	No	Filter: `all`, `api`, `guide`, `quickstart`, `help`, `component`

Example — search for authentication docs:

{
  "name": "searchDocs",
  "arguments": {
    "query": "authentication api key",
    "limit": 5
  }
}

getPage

Retrieve the full content of a specific page.

Parameter	Type	Required	Description
`slug`	string	Yes	Page path without `.mdx` (e.g., `guides/authentication`, `api-reference/pipeline/analyze`)

Example — fetch the quickstart page:

{
  "name": "getPage",
  "arguments": {
    "slug": "quickstart"
  }
}

Testing the endpoint

Verify the MCP server is reachable:

# List available tools
curl -X POST https://docs.boringmarketing.com/_mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'

# Search for content
curl -X POST https://docs.boringmarketing.com/_mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"searchDocs","arguments":{"query":"opportunities"}}}'

Useful page slugs

Quick reference for common pages your agent might want to fetch:

Slug	Content
`quickstart`	5-step zero-decision quickstart
`guides/authentication`	Auth, tiers, error codes
`guides/polling-runs`	Async 202 + poll pattern
`guides/content-production`	Brief → draft → publish flow
`guides/agent-integration`	Agent context block, endpoint table, workflows
`api-reference/overview`	Base URL, conventions, error format
`api-reference/pipeline/analyze`	Full pipeline endpoint
`api-reference/results/opportunities`	Opportunities endpoint + response shape
`api-reference/results/queue`	Action queue endpoint

Troubleshooting

Verify the endpoint is reachable:

curl https://docs.boringmarketing.com/_mcp

You should see a JSON response with server info. If not, the docs may not have been built yet.

The search index rebuilds with every deploy. If you recently added content, trigger a rebuild from the Jamdesk dashboard to update the index.

Make sure the slug matches the page path exactly, without the .mdx extension. For example, use guides/authentication not guides/authentication.mdx.