neuron-tool

Tool interface and registry for neuron agents

Overview

neuron-tool provides the Tool trait and ToolRegistry that operators use to expose callable functions to LLMs. Tools are described via JSON Schema, invoked with JSON arguments, and return JSON results.

ToolRegistry → serialized to tool_list → sent to model → model emits tool_call
             → ToolRegistry::call(name, args) → Tool::invoke → result back to model

Usage

[dependencies]
neuron-tool = "0.4"
serde_json = "1"

Defining a tool

use neuron_tool::{Tool, ToolError};
use serde_json::{json, Value};

pub struct UppercaseTool;

impl Tool for UppercaseTool {
    fn name(&self) -> &str { "uppercase" }

    fn description(&self) -> &str { "Convert text to uppercase" }

    fn input_schema(&self) -> Value {
        json!({
            "type": "object",
            "properties": {
                "text": { "type": "string", "description": "Text to convert" }
            },
            "required": ["text"]
        })
    }

    fn invoke(&self, input: Value) -> Result<Value, ToolError> {
        let text = input["text"].as_str().ok_or(ToolError::InvalidInput("missing text".into()))?;
        Ok(json!({ "result": text.to_uppercase() }))
    }
}

Registering tools

use neuron_tool::ToolRegistry;

let mut registry = ToolRegistry::new();
registry.register(UppercaseTool);

Part of the neuron workspace