Skip to main content
Make works well for visual routing after an agent has already collected data through Twexapi MCP.

Prerequisites

  • A Twexapi API key
  • A Make scenario with a Custom webhook
  • An MCP-capable agent connected to Twexapi MCP
  • A destination module such as Sheets, Airtable, Notion, Slack, or Data store

Scenario

  1. Add a Custom webhook module.
  2. Send Twexapi handoff JSON from your MCP-capable agent.
  3. Use Iterator on the tweets array.
  4. Add filters for topics, authors, or engagement.
  5. Write rows to Airtable, Google Sheets, Notion, or a database.

Agent prompt

Use Twexapi MCP to fetch trending tweets for country united-states, topic technology, content AI.
Return only JSON with route_used, country, topic, content, has_more, next_cursor, and tweets.
Each tweet must include tweet_id, author_username, text, created_at, like_count, reply_count, and retweet_count.

Iterator schema

{
  "tweet_id": "1803006263529541838",
  "author_username": "openai",
  "text": "Example tweet text",
  "created_at": "2026-07-02T10:00:00Z",
  "like_count": 10,
  "reply_count": 2,
  "retweet_count": 3
}

Direct REST continuation

For scheduled scenarios, call Twexapi REST directly from Make’s HTTP module and store next_cursor in a Make data store.

HTTP module setup

FieldValue
MethodPOST or GET from the selected endpoint
URLhttps://api.twexapi.io plus the relative path
HeaderAuthorization: Bearer YOUR_API_KEY
Body typeRaw JSON for POST requests

Router branches

BranchConditionAction
Empty resultstweets length is 0Stop quietly.
High engagementlike_count above thresholdSend Slack alert.
Continue pagehas_more is trueStore next_cursor.
Errorstatus is 401, 403, or 429Route to an ops notification.

Testing checklist

  • Run once with a small count, such as 5.
  • Confirm the Iterator receives one bundle per tweet.
  • Confirm tweet_id is used for dedupe before writing rows.
  • Confirm Make data store contains the latest next_cursor.