Last updated: 4 July 2026
Confluence
BetaSearch and retrieve Confluence pages, spaces, and documentation through your AI.
Who needs this: Teams using Confluence as their internal wiki who want AI to find and summarise documentation.
Example questions once connected
- Find the runbook for the M-Pesa reconciliation process
- What does the onboarding guide say about access requests?
- Get the contents of the API reference page
- Create a post-mortem page for the June 10 outage
Step-by-step setup
- Open the connector in your Msharti dashboard. Go to Connectors → Confluence → Connect. Msharti creates a unique OAuth link. Click Open in Browser.
- Log in with your Confluence account. A consent screen lists what Msharti can access. Review the permissions and click Accept or Continue.
- Confirm the connection in Msharti. You will be redirected back to the dashboard. The connector shows Connected with a green dot. Tokens are stored securely and refreshed automatically.
Tool reference
These are the tools the AI can call on this connector. Each tool fires automatically when you ask a relevant question.
search_pages() Full-text search across Confluence pages and spaces. Inputs query string, optional: space_key, limit
Try asking "Find all Confluence pages about the KRA integration"
Returns 3 pages: "KRA GavaConnect Setup" (IT/Integrations), "KRA API Notes" (Engineering/APIs), "Supplier Compliance Runbook" (Finance/Compliance).
get_page() Retrieve the full content of a specific Confluence page. Inputs page_title or page_id, optional: space_key
Try asking "Get the contents of the Supplier Compliance Runbook"
Returns Runbook: Supplier Compliance (Finance/Compliance). Steps: 1. Export supplier list from ERP... 2. Run KRA TCC check via Msharti... 3. Flag non-compliant...
create_page() Create a new page in a Confluence space. Inputs space_key, title, content, optional: parent_page_id
Try asking "Create a post-mortem page for the June 10 outage in the Engineering space"
Returns Page created: "Post-Mortem: June 10 Outage" in ENG space. URL: confluence.co/display/ENG/pm-jun10. Ready to edit.
Limits
- Same Atlassian OAuth token as Jira — connecting one connects both.
- Page content is extracted as plain text. Tables and embedded macros are included where parseable.
Need help? [email protected] — Nairobi-based, same time zone.