Classify synthetic requests while keeping authoritative review separate from AI output.
Workflow & automation · Documented · Documentation checked 2026-10-04
Make scenarios and Make AI Agents (New); connected model providers have their own terms.
Scenario limits vary by plan. AI Agents (New) is in open beta; custom AI providers require a paid plan.
An illustrative recipe. This website does not run the automation.
The goal and the right kind of system
Classify synthetic requests while keeping authoritative review separate from AI output.
A fixed scenario fits a known route. An agent would add unnecessary tool choice to a simple transformation.
Inputs and connected services
Inputs: stable request ID, original text, topic guide, expected fields, and controlled review state.
Services: Make and the selected model-provider path; any review store is a separately configured sandbox.
The steps
An illustrative sequence to adapt and test before using real data or enabling ongoing work.
- Run a manual sample scenario and inspect mapped bundles.
- Use an eligible AI step for a proposed classification.
- Validate output and route uncertainty to a draft queue.
- Require a separately controlled human approval field before any consequential module.
Outputs and human checkpoints
Outputs: private draft records and rejected / incomplete-input reasons.
Human checkpoints: review classification and exact action content. The model’s proposed approval cannot be used as the filter’s authoritative approval.
Exceptions, permissions, and stopping conditions
Permission boundary: no live sending in the initial scenario. Errors need an explicit route, and duplicate input needs reconciliation before a write.
Stop on invalid mapping, missing review, provider failure, or unsupported output. Inspect bundles before revising the prompt.
Adaptable prompt
Classify the fictional request from the approved topic guide. Return a draft topic and missing details only. Do not infer approval or consent. The fixed scenario will validate and route this output for human review.
Related reading
Scenarios, modules, and mapped data
Move records through a fixed scenario and inspect which values each module receives.
Filters, review gates, and error routes
Route only valid approved records to consequential actions and stop safely on failures.
Credits, provider billing, and run review
Track scenario usage and model-provider usage as different parts of operating cost.
Designing an AI workflow
Make a useful task repeatable with clear inputs, steps, outputs, exceptions, and human checkpoints.
Permissions and meaningful approvals
Decide what a system may read, what it may change, and when someone needs to authorize a concrete action.
Sources and verification
Documentation checked 2026-10-04. Examples are illustrative unless described otherwise.
Standard scenarios, AI scenarios, and agents · Make
Filtering bundles · Make
Error handling · Make
Credits and provider billing · Make