Bolt.new AI and Confluence integration: Step-by-Step Guide 2025

To integrate Bolt.new with Confluence, you don’t “connect” Bolt itself. What you actually do is: inside your Bolt.new project (Node/Next/Express/Python etc.), you call the real Confluence REST API using valid Atlassian authentication (either OAuth 2.0 via Atlassian Cloud or a Personal Access Token). Bolt is just your coding workspace — the integration is built like any normal backend API integration. Practically, this means: store your Confluence credentials in Bolt environment variables, call the Confluence REST endpoints from your server routes or actions, and return the results to your UI or to your AI agent. Nothing proprietary or magical exists — it’s plain Atlassian REST API usage.

 

What You Actually Have to Do

 

Confluence Cloud exposes a real REST API. Bolt.new lets you write backend code in a browser sandbox. So the integration is simply:

• Get authentication credentials (OAuth 2.0 recommended, Personal Access Token acceptable for internal use).
• Store those credentials in Bolt.new environment variables so your code can call Confluence securely.
• Write server code inside Bolt that performs REST calls to Confluence endpoints (example: search for pages, create pages, update content).
• Expose those calls via your own API routes for your UI or AI agent.
• Test inside Bolt sandbox, then migrate the code to your real environment.

 

Step-by-Step: Setting Up Confluence API Auth

 

Atlassian Cloud provides two real options:

• Personal Access Token (PAT): simplest; good for prototypes. Works like a password and uses Basic Auth.
• OAuth 2.0 (3LO): real production method; requires setting up an OAuth app in Atlassian Developer Console.

The simplest integration in Bolt.new is using a Personal Access Token.

• Create a Confluence Personal Access Token in your Atlassian cloud account.
• Store these in Bolt environment variables:
  CONFLUENCE\_EMAIL
  CONFLUENCE\_TOKEN
  CONFLUENCE_BASE_URL

Your base URL will look like:

https://your-domain.atlassian.net/wiki

 

Working Code Example (Node.js / Express inside Bolt.new)

 

This example fetches pages from Confluence to show how to perform API calls safely.

import express from "express"
import fetch from "node-fetch"

const app = express()

app.get("/confluence/pages", async (req, res) => {
  try {
    const url = `${process.env.CONFLUENCE_BASE_URL}/rest/api/content?type=page`

    // Basic authentication using email + PAT
    const auth = Buffer.from(
      `${process.env.CONFLUENCE_EMAIL}:${process.env.CONFLUENCE_TOKEN}`
    ).toString("base64")

    const response = await fetch(url, {
      method: "GET",
      headers: {
        "Authorization": `Basic ${auth}`,
        "Accept": "application/json"
      }
    })

    const data = await response.json()
    res.json(data)

  } catch (err) {
    console.error("Confluence API error:", err)
    res.status(500).json({ error: "Unable to fetch Confluence data" })
  }
})

app.listen(3000, () => {
  console.log("Server running on port 3000")
})

This code is valid and works exactly the same locally or inside Bolt.new’s backend runtime.

 

Creating a Page in Confluence

 

app.post("/confluence/create", express.json(), async (req, res) => {
  try {
    const url = `${process.env.CONFLUENCE_BASE_URL}/rest/api/content`

    const auth = Buffer.from(
      `${process.env.CONFLUENCE_EMAIL}:${process.env.CONFLUENCE_TOKEN}`
    ).toString("base64")

    const body = {
      type: "page",
      title: req.body.title,
      space: { key: req.body.spaceKey },
      body: {
        storage: {
          value: req.body.htmlContent,    // supports HTML storage format
          representation: "storage"
        }
      }
    }

    const response = await fetch(url, {
      method: "POST",
      headers: {
        "Authorization": `Basic ${auth}`,
        "Content-Type": "application/json"
      },
      body: JSON.stringify(body)
    })

    const data = await response.json()
    res.json(data)

  } catch (err) {
    console.error("Confluence API error:", err)
    res.status(500).json({ error: "Unable to create page" })
  }
})

 

How Bolt.new AI Fits In

 

Bolt AI doesn’t have a “Confluence plugin.” Instead:

• You write backend routes (like above) inside Bolt.
• You call those routes from your AI agent or UI.
• The AI can read/write Confluence through your server’s API, not directly.
• This keeps credentials safe and avoids exposing tokens to frontend or AI model.

This is the correct and real way to integrate AI features with Confluence.

 

Important Security Notes

 

• Never store PATs in client-side code.
• Always use server routes to proxy requests.
• Use OAuth 2.0 for production environments.
• Use Bolt environment variables during prototyping.