Confluence 集成

使用 LangChain JavaScript 集成 Confluence document loader。

兼容性

Only available on Node.js.

This covers how to load document objects from pages in a Confluence space.

凭证

• You’ll need to set up an access token and provide it along with your confluence username in order to authenticate the request
• You’ll also need the space key for the space containing the pages to load as documents. This can be found in the URL when navigating to your space e.g. https://example.atlassian.net/wiki/spaces/{SPACE_KEY}
• And you’ll need to install html-to-text to parse the pages into plain text

npm

npm install @langchain/community @langchain/core html-to-text

用法

import { ConfluencePagesLoader } from "@langchain/community/document_loaders/web/confluence";

const username = process.env.CONFLUENCE_USERNAME;
const accessToken = process.env.CONFLUENCE_ACCESS_TOKEN;
const personalAccessToken = process.env.CONFLUENCE_PAT;

if (username && accessToken) {
  const loader = new ConfluencePagesLoader({
    baseUrl: "https://example.atlassian.net/wiki",
    spaceKey: "~EXAMPLE362906de5d343d49dcdbae5dEXAMPLE",
    username,
    accessToken,
  });

  const documents = await loader.load();
  console.log(documents);
} else if (personalAccessToken) {
  const loader = new ConfluencePagesLoader({
    baseUrl: "https://example.atlassian.net/wiki",
    spaceKey: "~EXAMPLE362906de5d343d49dcdbae5dEXAMPLE",
    personalAccessToken,
  });
  const documents = await loader.load();
  console.log(documents);
} else {
  console.log(
    "You need either a username and access token, or a personal access token (PAT), to use this example."
  );
}

---

将这些文档连接到 Claude、VSCode 等工具，通过 MCP 获取实时答案。

在 GitHub 上编辑此页面 或 提交 issue。