Skip to main content

​
Overview

Atlassian is a one-click connector that connects Rootly AI to your Atlassian Cloud instance during investigations. A single OAuth flow connects both Jira and Confluence — Rootly AI can then look up related Jira issues to see if a bug was already reported, and search Confluence runbooks and architecture docs for how a system is supposed to work. Rootly AI can search and read Jira issues and Confluence pages. This connector passes through Atlassian’s available MCP tool catalog rather than enforcing a Rootly read-only tool list. The authorizing account’s Atlassian permissions determine which operations those tools can perform. Use an account scoped to the projects and actions you intend Rootly AI to reach, including during automatic investigations.

​
Before You Start

  • An Atlassian Cloud instance with permission to authorize third-party OAuth apps.
  • Read access to the Jira projects and Confluence spaces you want Rootly AI to be able to see.

​
Setup

Open AI SRE → Atlas → Connectors (AI & Agents → Connectors if your sidebar doesn’t have an AI SRE item) and click Connect on the Atlassian card. Enter a Connection name, then sign in to Atlassian and authorize Rootly AI for the site you want to connect. When the flow completes, the card flips to Connected.

​
What Rootly AI Can Read

Jira:
  • Issues — search across projects by keyword, assignee, status, or label.
  • Comments and history — the discussion thread on an issue.
  • Project context — project names, keys, and issue types.
Confluence:
  • Pages — search across spaces for runbooks, architecture docs, and past incident write-ups.
  • Page content — full body text of matched pages.
  • Space context — space names and structure so Rootly AI can distinguish “Team X’s runbooks” from “Team Y’s.”
Rootly AI reaches for Confluence when a question is about “how is this supposed to work?” — pairing it with a runbook-heavy Confluence and Sentry / observability data covers “what broke and how do we fix it?” investigations end-to-end.

​
During an Incident

“There’s already an open ticket for this exact error, filed last week. Here’s the link.” Rootly AI searches Jira for issues that match the current exception or error signature. If someone already reported it — bug ticket, customer complaint, tracked-work item — Rootly AI surfaces the link during the investigation, so the team doesn’t waste time re-triaging a known issue. On the Confluence side: “Here’s the runbook for this service.” When a responder asks how to handle a specific failure mode, Rootly AI searches Confluence for matching runbooks and cites the relevant one.

​
Best Practices

  • Authorize with an account that has broad read access. Rootly AI can only see the Jira projects and Confluence spaces the authorizing user has access to. If large chunks of your knowledge base are restricted, Rootly AI won’t be able to reason over them.
  • Keep runbook titles searchable. Confluence page titles are the primary signal Rootly AI uses to decide which page is relevant. Descriptive titles (“Payments API — on-call runbook”) beat generic ones (“Runbook”).
  • Link Jira issues to real components. Rootly AI can filter by project, but component-level context helps it surface the right related issue, not just any related one.

​
Troubleshooting

The authorizing Atlassian user doesn’t have access to those projects or spaces. Confirm the account has the right permissions, then disconnect and reconnect so the OAuth grant refreshes.
Confluence search ranks by title match and recency. If a stale page has a more matching title than the current one, Rootly AI may cite it. Rename or archive outdated pages so search surfaces the right ones.
If an Atlassian admin revokes the Rootly AI app from Atlassian’s side, Rootly’s card still shows Connected but every query fails. Reconnect from Rootly to reauthorize.

​
Frequently Asked Questions

No. A single Atlassian OAuth flow covers both. Which one actually returns data depends on the authorizing user’s permissions.
The one-click flow is built for Atlassian Cloud. If you’re on Data Center, contact support to see whether a Custom MCP connector makes sense for your setup.
The connector does not enforce a read-only tool list. Whether an Atlassian tool can create or change an issue depends on the connected account’s permissions and the tool Atlassian exposes. Review those permissions before enabling automatic AI SRE investigations. The existing Jira integration is a separate configuration for incident workflows.
Rootly does not maintain a persistent copy of Atlassian data — queries run at investigation time and the response is used in reasoning only. Query results do appear in the LLM traces Rootly logs for quality monitoring, as with every tool call Rootly AI makes. See Data Privacy for Rootly AI for the retention boundary.

Connectors Overview

All connectors and how they fit together.

Notion

The other knowledge-management connector.

Data Privacy for Rootly AI

What Rootly AI sees, retention, and model training controls.