---
title: Connect AI agents to Atlassian
slug: atlassian
integration: Atlassian
umbrella: Atlassian (Jira + Confluence)
category: pm
connectionType: oauth-rest
status: published
publishedAt: "2026-05-06T18:10:24.595Z"
lastUpdated: "2026-05-06T18:10:27.877Z"
canonical: "https://fidelic.ai/integrations/atlassian"
topSearchQueries:
  - "atlassian mcp"
  - "jira mcp"
  - "claude atlassian"
  - "jira ai integration"
---

# Connect AI agents to your Atlassian workspace

*AI agents that read your Jira tickets and Confluence pages, draft new ones, and post the result to Slack — sprint reviews, engineering briefs, QBR prep, and incident write-ups.*

## The connection

Fidelic agents connect to Atlassian through OAuth + the official Jira and [Confluence](https://www.atlassian.com/software/confluence) Cloud REST APIs. Your agent reads the projects and spaces the workspace admin explicitly shares with it — the same access a colleague would have, with scopes you control.

When the agent reads, it stays inside the projects you've shared. Nothing leaves your Atlassian workspace unless the agent posts the result to [Slack](https://slack.com/) or writes a draft into a destination space you authorized.

## Permissions requested

- Read projects, issues, and Confluence pages the workspace admin explicitly shares
- Create new pages and add comments in shared spaces
- Read user metadata for proper attribution on writes

Official docs: <https://developer.atlassian.com/cloud/jira/platform/rest/v3/intro/>

## Before you start

- An Atlassian Cloud workspace on any plan — Free, Standard, Premium, or Enterprise.
- Site admin role, or a workspace admin who can install third-party OAuth apps.
- A Slack workspace where the agent will live and post.
- A decision on which projects and spaces the agent should reach.
- About five minutes for the OAuth flow plus a thirty-minute background index build.

## How to connect

1. **Sign in to Fidelic and pick the agent.** From the Roster, choose the agent that needs [Atlassian](https://www.atlassian.com/) access. Each agent has an [Atlassian](https://www.atlassian.com/) connect button on its setup screen.
1. **Authorize **[**Atlassian**](https://www.atlassian.com/)**.** Click Connect [Atlassian](https://www.atlassian.com/). The OAuth dialog opens; sign in with your site admin account. The app appears in your site's Apps → OAuth credentials panel under the name Fidelic — [agent codename].
1. **Pick the projects and spaces the agent should reach.** Atlassian shows a permissions screen. Select the Jira projects and [Confluence](https://www.atlassian.com/software/confluence) spaces the agent will read or write to. You can change this any time from your Atlassian admin console.
1. **Confirm and finish.** The agent runs a test read against the first project you shared. When the test passes, your [Slack](https://slack.com/) channel shows a Connected to Atlassian message naming the agent.
1. **Try the first agent action.** Ask the agent a question in [Slack](https://slack.com/) — it will surface the answer with a Jira ticket or [Confluence](https://www.atlassian.com/software/confluence) page link.

## How you'll know it worked

Three signals confirm the connection is live.

- **Slack — **your channel shows a Connected to Atlassian message naming the agent and the projects it can reach.
- **Atlassian — **the Apps → OAuth credentials panel shows the Fidelic OAuth app with the projects and spaces you authorized.
- **First output — **the agent answers a Slack question with a citation that links back to a Jira ticket or Confluence page.

## Connection limits

Where the connection draws the line:

- **· **The agent only reads projects and spaces the admin explicitly shared. It does not auto-discover new content.
- **· **Jira automation rules and [Confluence](https://www.atlassian.com/software/confluence) macros run server-side; the agent reads their output, not the rules themselves.
- **· **Service Management (Jira Service Management) requires its own scope and is gated separately.

## Where the data goes

The connection runs inside your sandboxed [Anthropic](https://www.anthropic.com/) project. One project key per Fidelic customer; Fidelic infrastructure does not see your platform data.

Anything the agent reads stays inside your environment. The only data that leaves is what the agent posts to [Slack](https://slack.com/) or writes back to a destination you authorized — and you can revoke either at any time from the platform's connections panel.

---
Canonical: https://fidelic.ai/integrations/atlassian