DataToRAGDocs
Sign in

Confluence

Search with CQL, read and edit pages, read and add comments, look up attachments.

The Confluence tools let your AI assistant search space content, read and author pages, read and add comments, and look up an attachment's details.

Available operations

Tool Description
confluence_search Search content using CQL (Confluence Query Language)
confluence_list_pages List pages in a space with id, title, version, and link
confluence_get_page Get a page by ID with its version info. The body comes back as clean text by default; pass format: "storage" for the XHTML, which is what confluence_edit_page needs
confluence_create_page Create a new page. Content must be in XHTML storage format
confluence_edit_page Update an existing page. Version auto-increments if not provided
confluence_delete_page Delete a page by ID
confluence_get_comments Get all comments on a page with body content and version info
confluence_add_comment Add a plain-text comment to a page, optionally replying to an existing page comment. Inline comments cannot be replied to
confluence_get_attachment Get metadata for a page attachment by filename

Required scopes

Granular scopes, not classic ones: every Confluence tool except search calls the v2 API, which rejects a classic grant with scope does not match.

  • read:space:confluence, read:page:confluence
  • write:page:confluence, delete:page:confluence
  • read:comment:confluence, write:comment:confluence
  • read:attachment:confluence, search:confluence

Example prompts

  • "Read the sprint retro Confluence page and summarize the top action items"
  • "Search Confluence for our on-call runbook and pull out the escalation steps"
  • "List all pages in the Engineering space updated in the last week"
  • "Create a new Confluence page in the Product space with this week's roadmap review notes"
  • "Add a comment on the launch plan page listing the open questions for the design team"

Frequently asked questions

Why does Confluence need granular scopes rather than classic ones? #

Because of the API version. Every Confluence tool except search calls Atlassian's v2 API, which rejects a classic grant with a scope does not match error. Jira on the same connection still uses classic scopes.

What format does Confluence page content have to be in? #

XHTML storage format. The DataToRAG create-page tool takes content in Confluence's XHTML storage format. Reading a page returns clean text by default, along with version info; pass format storage to get the XHTML body, which is what an edit needs.

Do I have to supply a version number when editing a Confluence page? #

No. The Confluence edit tool auto-increments the version if you do not provide one.

Can it search Confluence, and how? #

Yes, with CQL. The Confluence search tool takes Confluence Query Language, and a separate tool lists the pages in a space with their id, title, version and link.

Can DataToRAG reply to an existing Confluence comment? #

Yes. The Confluence comment tool adds a comment to a page and can optionally reply to an existing page comment, and a companion tool reads every comment on a page with its body content and version info.

Try this on your own account

Connect Atlassian once and your assistant can run every tool on this page against your real data. Free to start, no card.

Get started free

By continuing you agree to our Terms of Service and Privacy Policy.

PreviousJira