Reported issues for Confluence Cloud

Pod holds 9 of 9 problems reported by people outside the maintainer team. Issues filed by the project's own owners, members and collaborators are excluded entirely — a maintainer's release checklist is not a warning to a prospective user.

Back to Confluence Cloud.

Most discussed

manage_confluence_page — get_versions returns stub message instead of version history

Description

Calling manage_confluence_page with operation: "get_versions" returns a stub message instead of version history:

Current version: 3

Full version history not yet implemented.

Suggestion

Either:

• Implement — the underlying Confluence API v2 exposes /wiki/api/v2/pages/{id}/versions with the data you'd need.
• Or remove get_versions from the operation enum until it's ready, so callers fail at parameter validation rather than getting an incomplete resp

Read the thread · 2026-05-12 · closed · 1 comment

Align CI/release process with groupware MCP server standard

Context

All four @aaronsb groupware MCP servers (confluence-cloud, jira-cloud, slack-mcp, google-workspace-mcp) negotiated a common CI/release configuration standard. This issue documents the consensus and the specific changes needed for confluence-cloud to converge.

Agreed Standard (all four projects)

1. Three CI workflows:
   • ci.yml — lint + test + build on PRs and pushes to main
   • release.yml — triggers on v* tag push, creates GitHub Release, builds and uploads .mcpb

Read the thread · 2026-04-11 · open · 1 comment

Version history, git-style diff between versions, and conflict detection on submit

Problem

manage_confluence_page with operation: "get_versions" currently returns:

Current version: 23
Full version history not yet implemented.

There is no way to list versions, read a historical version's content, or compare two versions. This bit hard today in a real workflow: two people editing the same SOW page. Session A pulled v19 into a scratchpad; a human published v20 from the browser; session A's submit landed as v21 built from v19, silently reverting v20. The human re

Read the thread · 2026-08-19 · open · 0 comments

Publish to npm by trusted publishing (OIDC) instead of a long-lived NPM_TOKEN

@aaronsb/confluence-cloud-mcp publishes from a long-lived NPM_TOKEN. google-workspace-mcp used to as well, and that token expired silently. This describes the OIDC path that replaced it, so this repo can follow the same convention.

This repo's current state: npm-publish.yml and release-mcpb.yml, publishing with NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}. It already sets id-token: write, passes --provenance, and has the prerelease dist-tag detection — so the change is small.

`s

Read the thread · 2026-08-17 · closed · 0 comments

manage_confluence_space — add a my_personal / whoami operation for current-user discovery

Description

Finding the calling user's personal space currently requires a roundabout chain:

1. search_confluence fulltext for the user's name (returns pages, not the space)
2. Pick a page authored by them, read its spaceId
3. manage_confluence_space operation: get to confirm the space identity

This costs 2–3 calls with non-obvious chaining, and only works if the user happens to have authored content recently.

Suggestion

Expose either:

• A new operation `manage_confluence_spa

Read the thread · 2026-05-12 · open · 0 comments

queue_confluence_operations — $N.field references don't support array indexing / JSONPath

Description

The queue's result-reference syntax ($N.field) works for flat top-level fields like pageId, but doesn't support array indexing or nested paths. This limits the queue to single-object handoffs.

What works

[
  {"tool": "edit_confluence_content",
   "args": {"operation": "submit", "scratchpadId": "..."}},
  {"tool": "manage_confluence_page",
   "args": {"operation": "add_labels", "pageId": "$0.pageId", "labels": ["..."]}}
]

$0.pageId resolves correctly. (Veri

Read the thread · 2026-05-12 · open · 0 comments

search_confluence — invalid CQL type returns JS error instead of a CQL diagnostic

Description

Submitting CQL that references an unsupported content type returns a confusing JS-layer error rather than a CQL-flavored diagnostic.

Repro

{
  "operation": "cql",
  "cql": "type = space AND space.type = \"personal\"",
  "limit": 5
}

Actual

Error: Cannot read properties of undefined (reading 'id')

Expected

Something like:

Error: CQL doesn't support type "space" (allowed: page, blogpost, comment, attachment)

Notes

CQL is documented a

Read the thread · 2026-05-12 · open · 0 comments

No working/documented syntax to embed an uploaded attachment inline in page body

Description

After uploading an image attachment via manage_confluence_media, there's no clearly documented or working way to inline-reference it in the page body through edit_confluence_content. Two of the three syntaxes I tried fail silently — no warning, no error, just an unusable page.

What I tried (all via edit_confluence_content → submit)

1. Markdown image syntax — brackets silently stripped

![Caption text](filename.png)

After submit and `pull_for_

Read the thread · 2026-05-12 · open · 0 comments

Most recent

manage_confluence_page — add_labels returns HTTP 405 Method Not Allowed

Description

The add_labels operation on manage_confluence_page consistently returns HTTP 405 from the Confluence API.

Repro

{
  "operation": "add_labels",
  "pageId": "<any valid pageId>",
  "labels": ["mcp-demo", "claude-authored"]
}

Actual

Error: Confluence API error 405: {"errors":[{"status":405,"code":"METHOD_NOT_ALLOWED","title":"org.springframework.web.HttpRequestMethodNotSupportedException: Request method 'POST' not supported","detail":null}]}

Exp

Read the thread · 2026-05-12 · closed · 0 comments

The remaining reports are on the project's issue tracker.