Algolia integration

What it does

The Algolia integration lets your agency drive a client's search experience from inside a workflow. Connect a client's Algolia application once and your workflows can run search queries against any index, push individual records into an index, partially update existing records by objectID, delete records, and submit batched add, update, and delete operations in a single request. A polling trigger walks an index on a fixed cadence so downstream workflows can react when new records appear without Algolia having to push to a webhook.

Connect an Algolia account

1. Open your workspace in TaskJuice and navigate to Connections.
2. Choose Algolia and click Connect.
3. In a new tab, open the Algolia API keys page signed in to the client's Algolia account.
4. Copy the Application ID from the top of the API keys page and paste it into TaskJuice. TaskJuice uses the Application ID both as the value of the X-Algolia-Application-Id header and as the host prefix on every Algolia request.
5. Decide which API key the connection should carry. Use the Admin API key when the workflow needs to add, update, or delete records. Use a Search-only API key when the workflow only needs to read. Scope the key to the indices the client wants TaskJuice to touch.
6. Copy the API key value, paste it into TaskJuice, and save the connection. TaskJuice sends it on every request as the X-Algolia-API-Key header.

To rotate or revoke the key, return to the Algolia API keys page, delete the entry, create a new key, and update the TaskJuice connection.

Triggers

• algolia/records-browsed polls the configured index by paging through the /indexes/{name}/browse endpoint on a fixed interval. Each cycle emits every record seen since the previous cursor watermark as a single activation, with the records wrapped in an items array.

Actions

• algolia/search-index runs a search query against a single index and returns matching hits, with optional filters, facets, and attribute projection. Paginates by page and hitsPerPage.
• algolia/save-object adds a new record to an index, or replaces an existing record when the supplied object carries an objectID.
• algolia/partial-update-object updates an existing record by objectID. Only the attributes you supply are written; the rest of the record is preserved.
• algolia/delete-object deletes a record from an index by objectID.
• algolia/batch-operations submits a batch of addObject, updateObject, partialUpdateObject, deleteObject, or clear operations against a single index in one request.

Known limitations

• The connected API key governs which actions succeed. A Search-only key cannot perform Save Object, Partial Update Object, Delete Object, or Batch Operations; configure an Admin or scoped write key for those workflows.
• Algolia does not push first-party webhooks for record changes. The Records Browsed trigger polls the browse endpoint instead, so latency between an upstream change and the next activation is bounded by the configured poll interval (minimum 300 seconds).
• Per-application rate limits, record-size limits, and operation quotas are governed by the client's Algolia plan. When Algolia returns a 429 the action surfaces as a retryable rate-limit error.
• Read actions route to the DSN read host ({appId}-dsn.algolia.net) and write actions route to the primary write host ({appId}.algolia.net). Per-region replica routing and multi-cluster write fan-out are not configurable from the connection.
• Each connection is scoped to a single Algolia application. To work with a second application, create a second connection.