COG Publish to Confluence Skill

When to Invoke

• User wants to publish a document to Confluence
• User says "publish to Confluence", "push to wiki", "send to Confluence", "confluence publish"
• User has a markdown file they want to share with the team on Confluence
• After generating a PRD, release notes, or other document that should go to the wiki

Agent Mode Awareness

Check agent_mode in 00-inbox/MY-PROFILE.md frontmatter:

• If agent_mode: team — no significant benefit for this skill (single sequential operation)
• If agent_mode: solo — standard execution

Command: /publish-to-confluence

Pre-Flight Check

1. Read 00-inbox/MY-INTEGRATIONS.md — Confluence MUST be listed under Active Integrations

   • If Confluence is not active: Inform the user and stop.

     Confluence is not in your active integrations.
     
     Would you like to:
     a) Set up Confluence integration (I'll add it to MY-INTEGRATIONS.md)
     b) Export the document in a Confluence-compatible format for manual upload
     

   • If Confluence is disabled: Skip silently per COG conventions. Suggest alternative publishing (HackMD, Notion) if active.

2. Get current timestamp: Run date '+%Y-%m-%d %H:%M' using Bash

---

Execution Strategy

Phase 1: Identify the Source Document

Determine what to publish:

Option A: User specifies a file path

Read the specified file from the vault.

Option B: User describes the document

Search for matching files:
1. Glob for likely matches in 04-projects/, 05-knowledge/, 01-daily/
2. Present candidates and let user choose

Option C: Just-generated document If this skill is invoked right after generating a PRD, release notes, or other document, use that document.

Phase 2: Configure Publishing Target

Ask the user (if not already provided):

Where should this be published in Confluence?

Required:
- Space key: [CUSTOMIZE: YOUR-SPACE-KEY] (or let me search for spaces)
- Parent page: [Page title or ID under which to nest this page]

Optional:
- Page title: [defaults to the document's H1 heading]
- Labels: [Confluence labels to add]
- Publish mode: "create new" or "update existing"

If the user doesn't know the space or parent page:

Use WebFetch to search Confluence:
GET /wiki/rest/api/space?limit=50
- List available spaces for the user to choose

GET /wiki/rest/api/content?spaceKey=[SPACE]&type=page&limit=25
- List pages in the space for parent page selection

Phase 3: Convert Markdown to Confluence Format

Convert the markdown document to Confluence storage format (XHTML):

Conversion rules:

Markdown	Confluence Storage Format
# Heading	<h1>Heading</h1>
## Heading	<h2>Heading</h2>
**bold**	<strong>bold</strong>
*italic*	<em>italic</em>
- list item	<ul><li>list item</li></ul>
1. list item	<ol><li>list item</li></ol>
[text](url)	<a href="url">text</a>
`code`	<code>code</code>
Code blocks	<ac:structured-macro ac:name="code"><ac:plain-text-body><![CDATA[...]]></ac:plain-text-body></ac:structured-macro>
Tables	<table><tbody><tr><th>...</th></tr><tr><td>...</td></tr></tbody></table>
> blockquote	<blockquote><p>quote</p></blockquote>
---	<hr />
- [ ] task	<ac:task-list><ac:task><ac:task-status>incomplete</ac:task-status><ac:task-body>task</ac:task-body></ac:task></ac:task-list>
- [x] task	Same with complete status
YAML frontmatter	Strip entirely (not displayed in Confluence)

Important conversion notes:

• Strip YAML frontmatter — it should not appear in the Confluence page
• Preserve table structure carefully
• Convert relative vault links to plain text (they won't work in Confluence)
• Handle nested lists properly
• Preserve emoji characters as-is

Phase 4: Preview and Approval Gate

CRITICAL: NEVER publish without explicit user approval.

Present a summary to the user:

Ready to publish to Confluence:

Document: [filename]
Title: [page title]
Space: [space key]
Parent Page: [parent page title]
Mode: [create new / update existing]
Labels: [labels]

Content preview (first 500 chars):
[preview text]

Estimated page size: [approximate word count]

Proceed with publishing? (yes/no)

Wait for explicit "yes" before proceeding.

Phase 5: Publish

Create New Page

Use WebFetch to POST to Confluence REST API:

POST [CUSTOMIZE: your-confluence-url]/wiki/rest/api/content

Headers:
  Content-Type: application/json
  Authorization: [from configured credentials]

Body:
{
  "type": "page",
  "title": "[page title]",
  "space": { "key": "[SPACE_KEY]" },
  "ancestors": [{ "id": "[PARENT_PAGE_ID]" }],
  "body": {
    "storage": {
      "value": "[converted XHTML content]",
      "representation": "storage"
    }
  },
  "metadata": {
    "labels": [
      { "prefix": "global", "name": "[label]" }
    ]
  }
}

Update Existing Page

First, get the current page to obtain the version number:
GET [CUSTOMIZE: your-confluence-url]/wiki/rest/api/content/[PAGE_ID]?expand=version

Then update:
PUT [CUSTOMIZE: your-confluence-url]/wiki/rest/api/content/[PAGE_ID]

Body:
{
  "type": "page",
  "title": "[page title]",
  "version": { "number": [current_version + 1] },
  "body": {
    "storage": {
      "value": "[converted XHTML content]",
      "representation": "storage"
    }
  }
}

Phase 6: Confirm and Update Vault

After successful publishing:

1. Confirm to the user:

   Published successfully!
   
   Page: [Title]
   URL: [Confluence page URL]
   Space: [Space Key]
   Version: [version number]
   

2. Update the source vault file (add published metadata to frontmatter):

   confluence_url: "[page URL]"
   confluence_page_id: "[page ID]"
   confluence_space: "[space key]"
   published_at: "[timestamp]"
   confluence_version: [version number]
   

3. Log the publication for future reference.

---

Re-Publishing (Update Flow)

If the vault file already has confluence_page_id in its frontmatter:

This document was previously published to Confluence:
  URL: [confluence_url]
  Last published: [published_at]

Would you like to:
a) Update the existing Confluence page (increment version)
b) Create a new page (separate copy)
c) Cancel

---

Fallback Behavior

Scenario	Behavior
Confluence not active	Stop and inform user; suggest alternative platforms
Confluence API fails	Save the converted XHTML to a file so user can manually paste it
Authentication failure	Inform user their Confluence credentials may need refreshing
Page already exists (same title)	Ask user: update existing, create with different title, or cancel
Very large document	Warn user about page size; suggest splitting into child pages
Complex markdown (unsupported elements)	Convert best-effort, warn about any elements that couldn't be converted

Error Handling

• API 403 (Forbidden): User may lack write permissions to the target space
• API 404 (Not Found): Space key or parent page ID may be incorrect
• API 409 (Conflict): Page was modified by someone else; fetch latest version and retry
• Conversion failures: Fall back to plain text for elements that can't be converted
• Rate limits: Retry with backoff
• Network errors: Save converted content locally so no work is lost