GibsonAI CLI Usage Examples

Building a Project End-to-End Using AI

1. Go to app.gibsonai.com

2. Chat with GibsonAI and create a new project.

3. Once complete, GibsonAI will email you the API key.

4. Configure the API key:

   gibson conf api::key [API key]

5. Import your project:

   gibson import api

✨ Magic, right?

Integrating a Model into Your Code

GibsonAI creates base-level SQLAlchemy models. To integrate them just do this:

from my_project.gibsonai.model.Module import ModuleBase

class MyModule(ModuleBase):
    pass

When you need to add custom business logic to the model, just modify your version of the model:

from my_project.gibsonai.model.Module import ModuleBase

class MyModule(ModuleBase):
    def my_custom_business_logic(self):
        return 1 + 1

We strongly suggest you do not refer to the base level SQLAlchemy models directly. By sub-classing you won't have to refactor your code if you decide to add custom code.

Integrating a Schema into Your Code

Currently, refer to the base-level schema directly.

Migrating Your Software from PHP to Python

1. Configure your datastore:

   gibson conf datastore::uri mysql+pymysql://[user]:[password]@[host]/[database name]

2. Turn on Dev Mode:

   gibson dev on

3. Import your MySQL schema:

   gibson import mysql

Well done! 70% of your code is written automatically. Customize the rest!

Asking GibsonAI Questions With Context

Use natural language queries:

gibson q [natural language request]

You can reference:

• Files: file://[full path]
• Python modules: py://[import]
• SQL entities: sql://[entity name]

Examples

gibson q format file:///Users/me/file.py for PEP8

gibson q code review py://user.modules.User

gibson q add nickname to sql://user

Using sql:// enables GibsonAI to:

• Create new entities
• Store them in last memory
• Let you generate models/schemas instantly

gibson code model user