Trappr
Docs/Platform Integrations/Make (Integromat)

Make (Integromat)

Protect your Make AI automation scenarios with Trappr — without building custom modules or writing code. Security monitoring integrates through the standard HTTP module.

How Make scenarios work

Make is fundamentally different from an agentic platform. There is no autonomous reasoning loop— a scenario is a fixed, ordered sequence of modules you define. Data flows left to right as structured bundles: each module receives the previous module's output, transforms it, and passes it forward. Every step is explicit; nothing happens unless you built it.

Because of this, the LLM call is exactly one HTTP module in the chain. Trappr intercepts at that specific module by changing its endpoint URL and Authorization header. Everything upstream (parsing, routing, filtering) and downstream (formatting, sending) stays exactly as it was.

Watch Email
trigger
→
Text Parser
bundle.text
→
HTTP: POST
TRAPPR HERE
gateway.trappr.net
→
Trappr Gateway
inspect · mask
→
OpenAI
completion
→
Send Reply
bundle.reply
Key insight: Make runs exactly one LLM call per scenario execution (unless you add a loop router). You change exactly one HTTP module — the URL and the Authorization header. Every other module in the scenario is untouched.

Prerequisites

  • A Make account with at least one scenario using an OpenAI or LLM call
  • A Trappr account with an app created and a Gateway key (cht_gw_*) ready
  • Your LLM provider key added in Trappr under Settings → Providers

Setup via HTTP module

1
Open your scenario and find the HTTP module calling OpenAI

In your Make scenario, click on the module that calls the OpenAI API (usually an HTTP → Make a request module or the OpenAI module).

2
Update the URL

Change the request URL from the OpenAI endpoint to the Trappr Gateway:

Text
# Before
https://api.openai.com/v1/chat/completions

# After
https://gateway.trappr.net/v1/chat/completions
3
Update the Authorization header

In the HTTP module headers, change the Authorization value from your OpenAI key to your Trappr Gateway key:

Text
Authorization: Bearer cht_gw_XXXX

Store the key in Make's Environment variables for security — don't hardcode it in the module.

4
Run the scenario and verify in Trappr

Execute the scenario once. In the Trappr dashboard under Gateway → Logs, confirm the request appeared with the correct model and token count.

Setup via OpenAI module

If your scenario uses Make's built-in OpenAI module:

1
Create a new OpenAI connection in Make

Go to Connections → Create a connection → OpenAI. In the API key field, enter your Trappr Gateway key. In the API URL field (if shown), enter https://gateway.trappr.net.

2
Switch your OpenAI module to use the Trappr connection

In your OpenAI module, switch the connection from your existing one to the Trappr-connected one. All other module settings remain the same.

The Make OpenAI module may not always expose a base URL field. If not, switch to the HTTP Request module — it gives you full control over the endpoint and headers.

Request body — unchanged

The Trappr Gateway accepts the exact same JSON body as the OpenAI API. Your existing request body — model, messages, temperature, max_tokens — requires no changes.

JSON
{
  "model": "gpt-4o",
  "messages": [
    {
      "role": "system",
      "content": "You are a customer support assistant."
    },
    {
      "role": "user",
      "content": "{{1.email_body}}"
    }
  ],
  "temperature": 0.3
}

Tamper detection for Make scenarios

Register your Make scenario as a Trappr agent to enable tamper detection. If any module, connection, or setting in the scenario is changed outside your normal deployment process, Trappr detects the deviation and alerts you immediately.

Go to Trappr → Agents → Register agent, select Make as the platform, and follow the on-screen instructions to link your scenario.

Tamper detection for Make is based on scenario structure snapshots. If you intentionally modify the scenario (e.g. a legitimate update), re-verify the agent fingerprint in the Trappr dashboard after saving.
← Previousn8nNext →LangChain / LangGraph