MCP Server
Heritage literature library via Model Context Protocol
Root.pk exposes an MCP server that gives AI assistants structured access to the heritage literature library. It uses Streamable HTTP transport — JSON-RPC 2.0 POST requests.
Connection
| Endpoint | https://root.pk/mcp |
| Transport | Streamable HTTP |
| Protocol | JSON-RPC 2.0 |
| Auth | None (public read-only) |
ChatGPT
Go to Settings → Connectors → Add, paste the endpoint URL, and save. ChatGPT will discover the tools automatically.
Claude
Add to your .mcp.json:
{
"mcpServers": {
"rootpk": {
"type": "http",
"url": "https://root.pk/mcp"
}
}
}
Tools
list-literature
List all available literature in the heritage library with their id, title, author, language and type.
No input parameters.
Response
| Field | Type | Description |
|---|---|---|
literature[].id |
integer |
Book ID |
literature[].title |
string |
Book title |
literature[].author |
string |
Author name |
literature[].description |
string |
Short description |
literature[].language |
string |
Language of the book |
literature[].type |
string |
Book type/category |
literature[].cover_url |
string |
Cover image URL |
search-literature
Search within a specific book for content related to a query. Returns the most relevant pages with their text and page URL.
Input
| Parameter | Type | Required | Description |
|---|---|---|---|
book_id |
integer |
yes | The ID of the book to search in. Use list-literature to find book IDs. |
query |
string |
yes | The search query. Must be written in the same language as the book — check the language field from list-literature and translate your query before calling this tool. |
limit |
integer |
no | Number of results to return. Defaults to 5. |
Response
| Field | Type | Description |
|---|---|---|
results[].page_number |
integer |
Page number within the book |
results[].distance |
number |
Cosine distance score (lower = more relevant) |
results[].markdown |
string |
Page content in markdown |
results[].url |
string |
Direct link to the page |