How to Integrate Retool with Confluence

Navigate to https://id.atlassian.com/manage-profile/security/api-tokens in your browser. This is Atlassian's central account management page and works for all Atlassian Cloud products including Confluence and Jira. Click Create API token. In the dialog, enter a descriptive label like Retool Integration and click Create. Copy the token immediately — Atlassian shows it only once and you cannot retrieve it later. The token is tied to your Atlassian account email address. In Confluence's Basic Auth scheme, you combine your email and this token to form the credentials: the Authorization header value is Basic followed by Base64-encoded email:token. Retool's Basic Auth option handles this encoding automatically — you just provide the username (email) and password (token) fields separately and Retool constructs the correct header. If you're building a shared Retool tool used by many team members, consider creating a dedicated Atlassian service account and generating the token from that account. This ensures the integration remains active even when individual users leave the organization. The service account needs Confluence space access or administrator privileges depending on what operations your Retool app needs.

In Retool, navigate to the Resources tab in the left navigation and click Add Resource. Select REST API from the connector list. In the configuration form: Name: Enter Confluence API (or something descriptive). Base URL: Enter https://YOUR_DOMAIN.atlassian.net/wiki/rest/api. Replace YOUR_DOMAIN with your actual Atlassian subdomain. For example, if your Confluence URL is https://acme.atlassian.net, the base URL is https://acme.atlassian.net/wiki/rest/api. Authentication: Select Basic Auth from the Auth type dropdown. In the Username field, enter your full Atlassian account email address. In the Password field, enter your API token (not your account password). Headers: Add a Content-Type header with value application/json. This is required for POST and PUT requests that send JSON bodies. Click Create Resource. Retool will save the configuration. To verify it works immediately, create a test query: GET /space with no parameters. If it returns a JSON object with a results array of your Confluence spaces, the connection is working correctly. For Confluence Server or Data Center (on-premise installations), the base URL format is https://YOUR_DOMAIN/rest/api and authentication uses your Confluence username and password (not email + API token).

CQL (Confluence Query Language) is Confluence's search query syntax, and it's the most powerful way to query content through the API. Use the /content/search endpoint with a cql query parameter. Create a query named searchPages. Set method to GET and path to /content/search. Add the following query parameters: - cql: your CQL query string (dynamic or static) - limit: {{ pagination.pageSize || 25 }} - start: {{ pagination.offset || 0 }} - expand: space,version,ancestors A CQL query to find all pages modified in the last 30 days across a specific space looks like: type = 'page' AND space.key = '{{ select_space.value }}' AND lastModified >= now("-30d") ORDER BY lastModified DESC For a free text search across all spaces: type = 'page' AND text ~ '{{ textInput_search.value }}' ORDER BY lastModified DESC For stale content auditing: type = 'page' AND lastModified < now("-90d") ORDER BY lastModified ASC The response includes a results array with pages, each having an id, title, type, space object, version object, and _links. Create a transformer to flatten this for display. Retool's Table component expects flat objects, not nested structures.

1// Transformer for Confluence search results
2const results = data.results || [];
3return results.map(page => ({
4  id: page.id,
5  title: page.title,
6  type: page.type,
7  space_key: page.space?.key || '',
8  space_name: page.space?.name || '',
9  version: page.version?.number || 1,
10  last_modified: page.version?.when ? new Date(page.version.when).toLocaleDateString() : '',
11  last_modifier: page.version?.by?.displayName || '',
12  url: `https://YOUR_DOMAIN.atlassian.net/wiki${page._links?.webui || ''}`,
13  status: page.status || 'current'
14}));

To let users select a space and browse its content, create a query that fetches all available Confluence spaces. Name it getSpaces, set method to GET and path to /space with these parameters: - limit: 50 - expand: description.plain,homepage - type: global (use 'personal' for personal spaces, or omit to get all) The response returns a results array where each space has a key (the short identifier like 'ENG' or 'MKTG'), name, type, and optional description and homepage objects. Set this query to Run on page load: yes in the query settings. Create a Select component (select_space) with data bound to {{ getSpaces.data.results }}, value mapped to key, and label mapped to name. This populates a space selector dropdown that drives your other queries. For Confluence instances with many spaces (100+), add pagination to the getSpaces query using the start parameter, or use CQL to filter spaces by type. You can also implement a search-as-you-type space finder using the /space?spaceKey={{ textInput.value }} endpoint with a debounced text input. If you need space statistics (page count, last activity), you'll need to combine the /space endpoint with additional queries since Confluence's space API doesn't include statistics in the base response.

1// Transformer for space dropdown
2const spaces = data.results || [];
3return spaces.map(space => ({
4  label: `${space.name} (${space.key})`,
5  value: space.key,
6  id: space.id,
7  type: space.type,
8  description: space.description?.plain?.value || 'No description',
9  homepage_id: space.homepage?.id || null
10}));

To display the full content of a selected page, create a getPageContent query. Set method to GET and path to /content/{{ table1.selectedRow.data.id }}. Add the expand parameter: body.view,version,space,ancestors,children.page. The body.view expansion returns the HTML-rendered content of the page. You can display this in a Retool Custom Component (iframe) or extract key metadata like word count and heading structure from it. For a lighter detail view without full HTML content, use body.storage instead of body.view — this returns Confluence's storage format XML which is easier to parse for metadata extraction. To show page history, add a getPageVersions query: GET /content/{{ table1.selectedRow.data.id }}/version with limit: 10 to show the 10 most recent versions. Bind this to a secondary Table in your detail panel showing version number, editor name, date, and a comparison link. For the admin panel layout, use a two-column Container: the left column shows the searchable page list (Table), and the right column shows the selected page's metadata — title, space, creator, created date, last modified, version count, and a button linking to the page in Confluence. Add a Labels section showing the page's labels fetched from /content/{id}/label.

1// Transformer for page detail view
2const page = data;
3const body = page.body?.view?.value || '';
4// Strip HTML tags to estimate word count
5const textContent = body.replace(/<[^>]+>/g, ' ').replace(/\s+/g, ' ').trim();
6const wordCount = textContent.split(' ').filter(w => w.length > 0).length;
7
8return {
9  id: page.id,
10  title: page.title,
11  space: page.space?.name || '',
12  version: page.version?.number || 1,
13  created: page.history?.createdDate ? new Date(page.history.createdDate).toLocaleDateString() : '',
14  last_modified: page.version?.when ? new Date(page.version.when).toLocaleDateString() : '',
15  last_editor: page.version?.by?.displayName || '',
16  word_count: wordCount,
17  status: page.status || 'current',
18  ancestors: (page.ancestors || []).map(a => a.title).join(' > '),
19  body_html: body
20};

Labels are key for organizing Confluence content and making it searchable. Add functionality to view and manage labels from your Retool panel. Create a getLabels query: GET /content/{{ table1.selectedRow.data.id }}/label. The response has a results array with label objects containing id, prefix, and name. To add a label: POST /content/{{ table1.selectedRow.data.id }}/label with body [{"prefix": "global", "name": "{{ textInput_label.value }}"}]. To remove a label: DELETE /content/{pageId}/label/{labelName}. For bulk operations, create a Retool Workflow that iterates over selected pages and applies or removes labels in batches. This is more reliable than in-app parallel queries for operating on 20+ pages at once. For page restrictions (making pages view-only or restricting editing), use the /content/{id}/restriction/byOperation endpoints. GET shows current restrictions; PUT replaces them entirely. Finally, implement a page hierarchy view using the ancestors array from the page detail query. Display ancestors as a breadcrumb trail (text labels connected by '>' separators) above the page title in your detail panel, so users always know where the page sits in the Confluence hierarchy.

1// Add label query body
2[
3  {
4    "prefix": "global",
5    "name": "{{ textInput_newLabel.value.toLowerCase().replace(/\\s+/g, '-') }}"
6  }
7]