AI and ML discovery

Last updated: Nov 11, 2026


The Discovery section of your organization shows an inventory of your assets. For each repository, Fluid Attacks finds the AI models that your code uses. For each model, you see its provider and the files where your code uses it.

Fluid Attacks finds the AI models in code that is written in these languages: C#, Go, Java, JavaScript, Kotlin, PHP, Python, Ruby, Swift, and TypeScript.

See the AI models of a repository

  1. In the left menu of your organization, click Discovery.
  2. In the Assets tab, click a repository.
  3. In the Details tab, go to the AI & Machine Learning section.

The section shows the number of AI models. Expand AI Models to see the first five models. For each model, the list shows its name, its provider, and the number of locations.

If the repository has no AI models, the section shows a message.

See all the AI models in a table

To see all the models, click View details. To see one model, click its number of locations. The AI Models pop-up window opens.

The table has these columns:

  • Model: The name of the AI model.
  • Provider: The company or service that supplies the model. This column is empty when the code does not name a provider.
  • Locations: The places where your code uses the model. For a file, a location has the path and the line, as in src/chat.py:29. When the line is not known, or for a directory, the location has only the path.

Use these options in the pop-up window:

  • Search: Type a model name or a location.
  • All providers: Select a provider to see only the models of that provider.
  • Download: Download the table as a CSV file.

Filter the assets by AI usage

In the Assets tab, the filters have an AI & Machine Learning group. Use it to find the assets that have AI Models, an MCP Server, or a Skill.

On this page