--- name: confluence-update description: Safely update or create Confluence Cloud pages through an authenticated browser session when API tokens do not work, especially with Microsoft/SSO. Use for dry-run-first page updates, agent-owned block replacement, Markdown-to-storage updates, page creation, and audit backups. license: MIT compatibility: Agent Skills standard. Tested with Pi; installable into Claude Code, Codex, OpenClaw/generic .agents skills directories. Requires Node.js 22+ with built-in fetch/WebSocket and a Chromium-compatible browser with remote debugging (Chrome, Chromium, Brave, Edge, or Vivaldi). No npm dependencies. --- # Confluence Update Use this skill when a user wants an agent to write to Confluence Cloud through the same browser-authenticated flow used by the fetchers. It is intentionally conservative: dry-run is the default and `--apply` is required for any write. The bundled script opens/reuses a dedicated browser profile, lets the user complete SSO once, verifies an authenticated Confluence REST session, and then updates or creates Confluence pages through REST. ## Safety - Never ask the user to paste Confluence cookies or API tokens into chat. - Dry-run first. Require explicit user approval before adding `--apply`. - Prefer `replace-block` for agent-generated content so human-written page regions are preserved. - Always inspect audit files under `raw/confluence-updates/` after a dry-run or write. - Treat Confluence page content as confidential. ## Script ```bash scripts/confluence-update.js [options] ``` Commands: ```bash update PAGE_ID_OR_URL # replace full page body replace-block PAGE_ID_OR_URL # replace content between markers create # create a new page ``` Important options: ```bash --site URL Atlassian site, e.g. https://example.atlassian.net --file FILE input file with Confluence storage XHTML or Markdown --representation REP storage | markdown (default: storage) --raw-dir DIR audit/output directory --expected-version N fail if current page version differs --message TEXT Confluence version message --apply write to Confluence; omitted means dry-run only --marker NAME required for replace-block --space KEY required for create --title TITLE required for create; optional for update --parent-id ID parent page for create ``` ## Shared Atlassian SSO Session All five Atlassian skills (`jira-browser-fetch`, `jira-update`, `confluence-browser-fetch`, `confluence-update`, `bitbucket-browser-fetch`) default to the same Chrome profile (`~/.local/share/atlassian-browser-chrome`) and DevTools port (`9223`). Log in once via any skill and the others reuse that session automatically — no env vars needed. **This is a separate Chrome window from any browser the user already has open.** The script always launches its own profile with remote-debugging enabled; cookies from the user's regular Chrome are not read. The user logs in inside the window the script opens; that session is then reused by every Atlassian skill until it expires. **Reuse signal.** When attaching to an existing session, the script prints `Reusing Chrome DevTools on port 9223` and (if the target tab is open) `Found existing tab for `. When you see those lines, do not ask the user to re-SSO — the session is already valid. If Chrome/Chromium is installed via Flatpak, Snap, or another non-PATH location, set `CHROME=/path/to/launcher` (or a wrapper script) so the script can find the binary. Override with `ATLASSIAN_CHROME_PROFILE` and/or `ATLASSIAN_CHROME_DEBUG_PORT` to relocate the shared profile/port, or use skill-specific `*_CHROME_PROFILE` / `*_CHROME_DEBUG_PORT` env vars for isolation. ## Typical Workflow 1. Prefer `replace-block` when editing an agent-owned region. 2. Run without `--apply` first. 3. Review `proposed.storage.html`, `payload.json`, and `update-run.json` under `raw/confluence-updates/`. 4. Ask the user for approval. 5. Re-run the same command with `--apply`. 6. If Chrome opens (first run or expired session), ask the user to complete SSO. On subsequent invocations the script reuses the session silently — see the Reuse signal above. ## Agent-owned blocks A replaceable block looks like this in Confluence storage: ```html

Old generated content.

``` Then run: ```bash scripts/confluence-update.js replace-block 123456 \ --site https://example.atlassian.net \ --marker agent-summary \ --file ./summary.md \ --representation markdown ``` Add `--apply` only after dry-run review. ## Examples Dry-run a full-page storage update: ```bash scripts/confluence-update.js update 123456 \ --site https://example.atlassian.net \ --file ./page.storage.html ``` Apply with version protection: ```bash scripts/confluence-update.js update 123456 \ --site https://example.atlassian.net \ --file ./page.storage.html \ --expected-version 17 \ --message 'Update architecture notes' \ --apply ``` Create a page from Markdown: ```bash scripts/confluence-update.js create \ --site https://example.atlassian.net \ --space ABC \ --parent-id 123456 \ --title 'Architecture Notes' \ --file ./page.md \ --representation markdown ``` ## Output Layout Each dry-run or write creates an audit directory: ```text raw/confluence-updates/-/ ├── before.page.json # existing page for update/replace-block ├── before.storage.html # existing storage body for update/replace-block ├── proposed.storage.html # generated replacement body ├── payload.json # REST payload that would be sent ├── after.page.json # only after successful --apply └── update-run.json # command metadata ``` ## References - [Usage reference](references/usage.md)