Skip to main content
The Confluence MCP server lets Corunner agents search, read, and write Confluence pages and spaces. This page covers how to connect it, configure usage guidance, verify capabilities, and manage the connection.

Connect with Confluence

Open the MCP Servers tab in the Corunner web app.

​
Server Details

​
Connect Confluence

1

Sign in to Corunner

Go to app.corunner.ai and sign in.
2

Open Integrations

Click Integrations in the left sidebar.
3

Select the MCP Servers tab

Choose the MCP Servers tab.
4

Find the Confluence card

Locate the Confluence card in the grid.
5

Turn on the Connect toggle

The Connect Confluence modal opens.
6

Verify the Server name

The prefilled name is Confluence. Rename it if you run multiple instances.
7

Verify the Server URL

Value: https://mcp.atlassian.com/v1/mcp/authv2. This field is managed by Corunner and is read-only.
8

Click Test URL

Confirm the endpoint format is valid and the server responds.
9

Select the authentication method

Prefilled: OAuth.
10

Click Advanced (Optional)

Expand the Advanced (Optional) panel. It contains the Usage guidance field, a free-text instruction that tells agents when and how to use this server. Corunner appends what you enter here to every tool description sent to the agent, so it directly influences when the agent picks Confluence over another source. Leave it empty now and add it later from Edit configuration if you prefer.
11

Enter Usage guidance

In the Usage guidance field (placeholder: Tell agents when and how to use this server…), describe when this server should be used, which spaces or pages to prefer, what needs human approval, and what must not be accessed. Keep it short and imperative. Max 2000 characters. See the next section for a copyable example.
12

Click Connect with Confluence

This initiates the OAuth authorization flow.
13

Authorize with Atlassian

Corunner redirects the current tab to the Atlassian authorization page.
14

Approve permissions

Review and approve the requested permissions in the Atlassian consent screen.
15

Return to Corunner

After approval, the tab returns to Corunner automatically.
16

Confirm connection

Confirm a green Connected badge and an enabled Connected toggle on the Confluence card.

​
Configure Usage Guidance

Usage guidance tells Corunner when, how, and what boundaries to apply when agents use this server. The six purposes are:
  1. When to use it
  2. What to do with it
  3. Which resources are permitted
  4. Which resources require approval
  5. What must not be accessed
  6. How to present results
Never enter secrets, tokens, or credentials into the Usage guidance field. It is stored as plain text and visible to all workspace members.
Copy and adapt the following example:

​
Verify the Connection

After connecting, the Confluence card shows:
  • A green Connected status pill next to the name
  • An enabled Connected toggle
  • Connected to <workspace> with a relative timestamp (e.g., “2 minutes ago”)
  • A <N> capabilities button that expands into the capability list

​
Capabilities

The Confluence server exposes the following tools. Confluence exposes both read and write tools from the connected server.
  • search pages
  • read spaces
  • connected server write tools
Click the capability count or its chevron on the connected card to view the current tool list after connecting.

​
Check Connection Health

Click the Test connection icon (Wi-Fi glyph) on the Confluence card to run a live probe. A green check means the endpoint is reachable and credentials are valid. A red warning means the probe failed. If the test fails:
  • Open Edit configuration to refresh your OAuth credentials or verify the Server URL

​
Edit Configuration

Click the Edit configuration icon (settings glyph) on the Confluence card. Admins can update:
  • Server name
  • Server URL (read-only for Corunner-managed endpoints)
  • Authentication method and credentials
  • Custom headers (non-OAuth only, under Advanced)
  • Usage guidance
Create, edit, disconnect, and observation-consent changes are admin-only. Members can view status and capabilities but cannot modify settings.
Click the Observation consent icon (brain glyph) to open the consent drawer and configure what Confluence resources Corunner can observe for organizational memory. Observation is separate from tool connection: it controls whether Corunner passively reads resources into memory, while capabilities (tools agents can call) are unaffected.
  • Mode: Disabled / Shadow / Active
  • Sensitivity: Standard / Restricted / Excluded
  • Excluded scopes cannot be enabled
Observation is separate from connecting a tool. Only resources listed here are eligible for passive intelligence, and excluded scopes cannot be enabled.
The precise effect on Memory & Learnings for advanced observation modes should be confirmed with the Corunner product team if you plan to depend on it in policy.

​
Disable or Disconnect

Disconnecting deletes stored credentials for this connection. Provider-side OAuth tokens should also be revoked from the provider account if you want to eliminate them entirely.

​
Connect Confluence

Connect with Confluence

Ready to go? Open the MCP Servers tab and turn on the Confluence Connect toggle.

Direct Jira & Confluence

Set up a direct connection to Jira and Confluence without MCP.

Integrations Overview

Browse all available integrations and MCP servers.

Memory & Learnings

Learn how Corunner builds organizational memory from connected sources.