Tools

Tools are APIs or custom functions the agent can use to perform actions or fetch data. Once you equip an agent with a tool, it will use it automatically when it needs to based on context.

You can turn any API or function into a tool the agent can use, you just need to define the schema of the inputs the tool expects to receive from the LLM.

​

Validation

All data is type-safe and VALIDATED… so you can build reliable LLM-powered applications and your tools never receives unexpected inputs.

​

Errors recovery

Validation errors and tool execution errors are handled intelligently by Scoopika… how? well it’s based on the type of the error.

If an input is missing we try to figure it out or ask the user to provide it, if a value is invalid we try to solve the issue, and if the tool execution failed we investigate the issue and maybe try later or just tell the user about it.

All of that is done for you under the hood, from the tool execution, to the validation, type-safety, and errors recovery.

​

Types of tools

There are different types of tools that agents can call in Scoopika:

​

API tools

These tools are defined from the platform with no code, and allows the agent to send HTTP requests to external APIs.

The response of the HTTP request is returned back to the agent to process it. Can be used to perform actions or fetch data.

​

Custom functions (server)

Custom JS functions defined and executed on the server-side. you can turn any function into a tool the agent can use.

The result of the function is returned back to the agent to process it. Can be used to perform actions or fetch data.

​

Client-side actions

Custom JS functions defined and executed safely on the client-side by the agent in real-time (no eval or anything unsafe is used).

The result of the function is NOT returned back to the agent. Can only be used to perform actions on the client-side like clicking buttons or generating UI.