Rovo natively connects to 60+ applications as of August 2026 — but Microsoft Planner, OneNote, and QuickBooks aren’t among them, despite all three sitting next to Jira in plenty of real orgs. Atlassian’s own connectors page lists the official set across project management, collaboration, dev, and communication tools; the newer Rovo MCP Server covers a different job — piping Atlassian data out to external AI clients instead of pulling other tools’ data in. This page maps both, then covers what’s genuinely missing.

What connectors does Rovo support natively?

Per Atlassian’s own connectors page (checked live August 2026), the official list runs to over 60 applications across five categories:

  • Project management — Aha!, Airtable, Asana, ClickUp, Jira, Jira Data Center, Jira Product Discovery, Jira Service Management (+ Data Center), Monday, ServiceNow (+ Tickets), Smartsheet, Trello.
  • Collaboration — Adobe Sign, Adobe XD, Box, Canva, Confluence (+ Data Center), Custom Websites, Docusign, Dovetail, Dropbox, Figma, Google Drive, Lucid, Miro, Mural, Notion, OneDrive, SharePoint.
  • Development — Azure DevOps, Bitbucket, GitHub, GitLab, LaunchDarkly, Sentry.
  • Communication — Gmail, Google Calendar, Microsoft Teams, Outlook Calendar, Outlook Mail, Slack, WebEx.
  • Additional tools — Amplitude, Atlassian Home, Compass, HubSpot, Loom, Pipedrive, Power BI, Salesforce, Stripe, Todoist, Workday, Zendesk, Zeplin.

Your own Jira and Confluence don’t need a connector at all — they’re the native core of the Teamwork Graph, per Atlassian’s own framing of Rovo as built directly on years of Jira history and Confluence content. The explicit “Jira,” “Jira Data Center,” and “Confluence Data Center” entries on the connector list cover a narrower case: pulling in additional Jira/Confluence instances beyond your home site, or an on-prem Data Center deployment that isn’t part of the cloud graph by default.

For the DevOps-specific slice of this list — GitHub, GitLab, Azure DevOps, and how Rovo Dev and Bitbucket fit around it — the deeper breakdown lives in Rovo DevOps integration, explained; this page stays at the full-landscape level.

What’s the difference between a “synced” and a “direct” connector?

Not all connectors work the same way underneath, and the distinction affects both search quality and your data footprint. Per Atlassian’s own docs:

  • Synced connectors index and store your workspace content inside Atlassian’s systems using a service or admin account. This gives the best search and chat quality, because the content is fully indexed rather than fetched on demand — but every object synced counts against your org’s indexed-objects allowance.
  • Direct connectors — currently Outlook Mail, Gmail, and Slack — query the provider’s own search API live, every time, and store nothing on Atlassian’s side. Nothing counts toward the indexed-objects allowance, but search quality is described as medium rather than the synced tier’s comprehensive coverage.

The practical read: if data residency matters more than search depth for a given tool, that’s the case for a direct connector where one’s available. Otherwise, synced is the default and the stronger search experience.

How do you enable a Rovo connector?

It’s an admin action, not a user toggle — done at admin.atlassian.com by an org admin, per Atlassian’s connector management docs. The admin picks the connector, authorizes it against the target tool (usually an OAuth consent flow or a service account), and it starts feeding the Teamwork Graph.

One detail worth knowing before you enable anything for a team: Rovo doesn’t grant new access. It surfaces content through the permissions a user already has in the connected tool — someone who can’t see a Confluence space or a Slack channel today still won’t see it through Rovo Search after the connector’s live.

What is the Rovo MCP Server, and how is it different from a connector?

A connector’s job is bringing outside data into Rovo’s Teamwork Graph so Rovo’s own Search, Chat, and Agents can use it. The Rovo MCP Server runs the opposite direction: it exposes that same graph out to external AI clients — Claude, Cursor, ChatGPT — over the Model Context Protocol, so an agent you’re already running can read your Jira and Confluence context without going through Rovo’s chat UI at all.

If your team’s actual need is “I want Claude Code to see our Jira tickets,” that’s an MCP question, not a connector question — the full setup, auth model, and how it compares to the community’s mcp-atlassian server is covered in Rovo MCP Server, explained.

One thing both share: per Atlassian’s usage-allowance docs, plain search across connected tools doesn’t burn Rovo credits — only Chat and agent runs do, at 10 credits per request against your plan’s pooled monthly allowance (25/user on Standard, 70 on Premium, 150 on Enterprise). Connecting more tools costs nothing extra to try; the meter only starts when an agent actually acts on what it finds.

What’s NOT covered by Rovo’s connectors?

This is the gap that matters if you’re evaluating Rovo for a real stack rather than reading the marketing page. Checked against Atlassian’s own list (August 2026), three tools that show up constantly next to Jira in real orgs are simply absent:

  • Microsoft Planner — no connector, despite Planner being a Microsoft 365 staple that plenty of teams run alongside Jira for lighter-weight task tracking.
  • OneNote — also missing, despite living in the same Microsoft 365 suite as the SharePoint and OneDrive connectors Rovo does ship.
  • QuickBooks — absent from the “Additional tools” category entirely, even though HubSpot, Salesforce, and Stripe are covered.

None of these are obscure — they’re mainstream tools that happen to sit outside Atlassian’s own connector roadmap. If your team’s context lives partly in one of them, Rovo’s Teamwork Graph simply won’t see it until someone builds the bridge.

Can you build a custom Rovo connector?

Yes, but it’s a development project, not a settings page. Per discussion on Atlassian’s own developer community, there are two real routes today:

  1. The custom website crawler — read-only, fine for basic web content, not built for a structured app’s data model.
  2. A Forge app implementing Rovo Connector modules — the real path for indexing a SaaS tool properly, which means writing and deploying an actual Forge app against that tool’s API.

Atlassian has floated a simpler content-ingestion API for basic indexing, but as of this list it’s roadmap, not shipped. So connecting something like Planner, OneNote, or QuickBooks to your Atlassian workflow today means either building that Forge integration yourself or reaching for a purpose-built connector that already exists outside Rovo’s own list — which is exactly the gap our own products sit in: MS Planner to Jira, OneNote Reports for Jira, and QuickBooks Timesheet Sync for Jira & Tempo all bridge tools that never made Atlassian’s own connector list.

If the connector you need doesn’t exist yet, we build them.