M
Manifest•4w ago
aitorroma

Exploring Manifest's Potential: Typebot API Use Case & Feature Suggestions

First of all, thank you very much for Manifest. I believe it has great potential as an application, making it much easier to develop new projects. I have been testing the platform and wanted to explore a use case with Typebot, using it as an API to integrate it with a chatbot. My goal is simply to assess its potential in this context. I have some recommendations that could enhance the experience: - Support for auto-increment fields in orders. - Ability to customize the body of webhooks. - Capability to perform automatic calculations for totals. - Option to create custom views, similar to designs in CouchDB. - A cron-type field that would allow executing webhooks based on a scheduled time. This would be particularly useful for reminders. Below, I am sharing an example of the use case I am trying to implement: 
name: Order Management App
entities:
  Customer:
    seedCount: 200
    properties:
      - { name: name, type: string }
      - { name: email, type: string }
      - { name: phone, type: string }
      - { name: address, type: text }
    
  Category:
    properties:
      - { name: name, type: string }

  Product:
    properties:
      - { name: name, type: string }
      - { name: current_price, type: money, options: { currency: 'EUR' } }
      - { name: stock, type: number }
      - { name: description, type: text }
      - { name: image, type: image }
      - { name: active, type: boolean }
    belongsToMany:
      - Category

  Order:
    properties:
      - { name: date, type: timestamp }
      - { name: total, type: number }
      - { name: status, type: choice, options: { values: [pending, processing, shipped, delivered, canceled] } }
    belongsTo:
      - Customer
    belongsToMany:
      - OrderItem

  OrderItem:
    properties:
      - { name: quantity, type: number }
    belongsTo:
      - Order
      - Product
name: Order Management App
entities:
  Customer:
    seedCount: 200
    properties:
      - { name: name, type: string }
      - { name: email, type: string }
      - { name: phone, type: string }
      - { name: address, type: text }
    
  Category:
    properties:
      - { name: name, type: string }

  Product:
    properties:
      - { name: name, type: string }
      - { name: current_price, type: money, options: { currency: 'EUR' } }
      - { name: stock, type: number }
      - { name: description, type: text }
      - { name: image, type: image }
      - { name: active, type: boolean }
    belongsToMany:
      - Category

  Order:
    properties:
      - { name: date, type: timestamp }
      - { name: total, type: number }
      - { name: status, type: choice, options: { values: [pending, processing, shipped, delivered, canceled] } }
    belongsTo:
      - Customer
    belongsToMany:
      - OrderItem

  OrderItem:
    properties:
      - { name: quantity, type: number }
    belongsTo:
      - Order
      - Product
Thanks again for this great tool!
2 Replies
brunobuddy
brunobuddy•4w ago
Hello @aitorroma thank you very much for your message !! The features requested are really nice. Some of them are on our roadmap and some of them are already developped. We have endpoints now that you can use for performing calculations or custom logic in JavaScript: https://manifest.build/docs/endpoints Let me know if you have any question !
Endpoints | Manifest Docs
A custom endpoint is a user-defined API route that executes specific logic on the backend. Unlike built-in routes, custom endpoints allow you to control how data is processed, retrieved, or modified in response to client requests.
From An unknown user
From An unknown user
aitorroma
aitorromaOP•4w ago
Ohh! Thanks! I found your project interesting because it simplifies and makes API stacks much more portable compared to solutions like PocketBase or Strapi.
For example, for people like me who have automation and AI communities, it is much easier to share a YML file than to manually develop all the tables or use SQL code.
Additionally, since the project is well-structured and in plain text, it would be possible to create GPTs that generate stacks for specific applications.
I think it would be great to quickly implement it with Docker and allow uploading YML files from the back office to generate the entire API structure directly from there.
It would also be useful to download the OpenAPI specification, as it could enable easy interface generation using AI and OpenAPI specifications. @brunobuddy I've also sent you a private message in case you're interested in collaborating on an interview for my newsletter.

Did you find this page helpful?