> ## Documentation Index
> Fetch the complete documentation index at: https://linkly.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Using Linkly AI in QwenPaw

> Connect your Linkly AI local knowledge base to the QwenPaw agent, letting AI search and read your private documents

## What You Can Do

<AccordionGroup>
  <Accordion title="Search local documents during task execution">
    QwenPaw supports extending tool capabilities through MCP. After connecting
    Linkly AI, the agent can search your local documents at any time while
    planning and executing tasks, combining local knowledge with online
    information to improve output quality.
  </Accordion>

  <Accordion title="Use local materials in planning mode">
    In QwenPaw's planning mode, let the agent first retrieve relevant documents
    through Linkly AI, then create a detailed execution plan — ensuring the plan
    is built on your real private materials.
  </Accordion>

  <Accordion title="Share one knowledge base across multiple agents">
    QwenPaw supports configuring multiple independent agents. After connecting
    Linkly AI, each agent can access the same local documents without setting up
    separate knowledge sources for each.
  </Accordion>
</AccordionGroup>

## Prerequisites

* [QwenPaw](https://qwenpaw.agentscope.io/downloads) installed
* Linkly AI Desktop running with the MCP service active

## Configuration Steps

<Tabs>
  <Tab title="Local MCP HTTP (Recommended)" icon="globe">
    <Steps>
      <Step title="Open MCP configuration">
        Open the QwenPaw console, go to **Agent → MCP**, and click **+ Create**.
      </Step>

      <Step title="Paste the configuration and create">
        Paste the following JSON in the dialog, then click **Create**:

        ```json theme={null}
        {
          "mcpServers": {
            "linkly-ai": {
              "transport": "streamable_http",
              "url": "http://127.0.0.1:60606/mcp"
            }
          }
        }
        ```
      </Step>
    </Steps>
  </Tab>

  <Tab title="Remote MCP" icon="cloud">
    Complete [remote access configuration](/en/use-in-remote) first, with the
    desktop tunnel showing <Badge stroke color="green">● Connected</Badge>.

    <Steps>
      <Step title="Open MCP configuration">
        Open the QwenPaw console, go to **Agent → MCP**, and click **+ Create**.
      </Step>

      <Step title="Paste the remote configuration and create">
        Paste the following JSON in the dialog, then click **Create**:

        ```json theme={null}
        {
          "mcpServers": {
            "linkly-ai": {
              "transport": "streamable_http",
              "url": "https://mcp.linkly.ai/mcp"
            }
          }
        }
        ```
      </Step>
    </Steps>
  </Tab>

  <Tab title="CLI Stdio (Fallback)" icon="terminal">
    [Install Linkly AI CLI](/en/use-cli) first, and ensure Node.js 18+ is
    installed locally.

    <Steps>
      <Step title="Open MCP configuration">
        Open the QwenPaw console, go to **Agent → MCP**, and click **+ Create**.
      </Step>

      <Step title="Paste the CLI configuration and create">
        Paste the following JSON in the dialog, then click **Create**:

        ```json theme={null}
        {
          "mcpServers": {
            "linkly-ai": {
              "command": "linkly",
              "args": ["mcp"]
            }
          }
        }
        ```
      </Step>
    </Steps>
  </Tab>
</Tabs>

## Getting Started

Once configured, ask questions directly in a QwenPaw conversation and the agent
will automatically call Linkly AI tools to search and read your local documents.

Examples:

* "Search my documents for content about XX"
* "Check if my local notes have anything about YY and summarize it"

## FAQ

<AccordionGroup>
  <Accordion title="Tools not working after creation">
    Go to **Agent → MCP** and confirm the enable toggle for the linkly-ai client
    is on. Also check that Linkly AI Desktop is running and the MCP service
    shows Running.
  </Accordion>

  <Accordion title="CLI mode: command not found">
    Confirm that Linkly AI CLI is correctly installed (verify with `linkly     --version`) and that Node.js version ≥ 18 (`node --version`).
  </Accordion>

  <Accordion title="Can I import multiple MCP configurations at once">
    Yes. QwenPaw supports configuring multiple MCP clients in the `mcpServers`
    object of a single JSON — one click creates them all at once.
  </Accordion>
</AccordionGroup>