# The paste holds the whole tool list

**When a tools/list answer carries nextCursor it is only one page; call tools/list again with that cursor until none comes back, and check all pages together.**

Check `spec-pagination-partial` · MCP specification 2026-07-28 · default severity: advice · rubric 2.1.0. Last verified against MCP specification 2026-07-28: Tools on 2026-11-18.

## What fails

The answer carries a nextCursor, so it is one page. Tools on later pages were not checked, and the duplicate-name checks only see this page.

## The rule

> This operation supports pagination and caching.

Source: [MCP specification 2026-07-28: Tools](https://modelcontextprotocol.io/specification/2026-07-28/server/tools#listing-tools)

## How to fix it

Fetch every page and paste them together as a JSON array of the answers; the checker merges them.

1. Call tools/list, then repeat with params.cursor set to nextCursor until it is absent.
2. Paste the answers as a JSON array; the checker merges them.
3. Or let the client SDK collect all pages for you, as in the code.

### TypeScript SDK

```ts
import { McpServer } from '@modelcontextprotocol/server';
import * as z from 'zod';

const server = new McpServer({ name: 'issues', version: '1.0.0' });

server.registerTool(
  'get_issue',
  {
    title: 'Get issue',
    description: 'Return one issue by ID. Use when the user names an issue.',
    inputSchema: z.strictObject({
      issue_id: z.string().describe('Issue ID, for example ISS-42'),
    }),
    annotations: { readOnlyHint: true, destructiveHint: false, openWorldHint: false },
  },
  async ({ issue_id }) => ({ content: [{ type: 'text', text: 'Done' }] }),
);

// Client side: collect every page of tools/list.
// let cursor; const tools = [];
// do { const page = await client.listTools(cursor ? { cursor } : {}); tools.push(...page.tools); cursor = page.nextCursor; } while (cursor);
```

### Python SDK

```python
from typing import Annotated

from pydantic import Field
from mcp.server.mcpserver import MCPServer
from mcp.types import ToolAnnotations

server = MCPServer("issues")

@server.tool(
    title="Get issue",
    annotations=ToolAnnotations(readOnlyHint=True, destructiveHint=False, openWorldHint=False),
)
def get_issue(
    issue_id: Annotated[str, Field(description="Issue ID, for example ISS-42")],
) -> str:
    """Return one issue by ID. Use when the user names an issue."""
    return "Done"

# Client side: collect every page of tools/list.
# tools, cursor = [], None
# while True:
#     page = await session.list_tools(cursor=cursor)
#     tools += page.tools
#     cursor = page.next_cursor
#     if not cursor:
#         break
```

## Questions

### Do clients fetch all pages?

Conforming clients do. A hand-rolled script or a quick curl often stops at the first page, which is why the checker flags a partial paste.

### Why does it matter for checking?

Tools on later pages are not graded, and duplicate names split across pages are missed.

## Related checks

- [Tool names are unique within the server](https://isokron.ai/mcp-check/checks/spec-name-unique)
- [The tools/list result carries ttlMs and cacheScope](https://isokron.ai/mcp-check/checks/spec-envelope-cache)
- [Every list entry is a tool object with a name](https://isokron.ai/mcp-check/checks/spec-tool-shape)
- [MCP specification 2026-07-28: tools/list rules](https://isokron.ai/mcp-check/checks/protocol)

Counts toward: MCP specification 2026-07-28.

Rubric: https://isokron.ai/mcp-check/rubric

Check your tool list: https://isokron.ai/mcp-check
