Struct CodeExecutionTool 

pub struct CodeExecutionTool { /* private fields */ }

Code execution tool.

This tool allows agents to execute code in a sandboxed environment. It requires integration with an external code execution service.

Safety

Code execution is inherently dangerous. This tool should:

• Always run in a sandboxed environment
• Have strict resource limits
• Only be used with trusted agents

Example

use serdes_ai_tools::builtin::{CodeExecutionTool, CodeExecutionConfig, ProgrammingLanguage};

let tool = CodeExecutionTool::with_config(
    CodeExecutionConfig::new()
        .timeout_secs(10)
        .allowed_languages(vec![ProgrammingLanguage::Python])
);

Implementations

impl CodeExecutionTool

pub fn new() -> Self

Create a new code execution tool with default config.

pub fn with_config(config: CodeExecutionConfig) -> Self

Create with a specific config.

Trait Implementations

impl Debug for CodeExecutionTool

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CodeExecutionTool

fn default() -> Self

Returns the “default value” for a type.

impl<Deps: Send + Sync> Tool<Deps> for CodeExecutionTool

fn definition(&self) -> ToolDefinition

Get the tool’s definition.

fn call<'life0, 'life1, 'async_trait>( &'life0 self, _ctx: &'life1 RunContext<Deps>, args: JsonValue, ) -> Pin<Box<dyn Future<Output = ToolResult> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Execute the tool with given arguments.

fn max_retries(&self) -> Option<u32>

Maximum retries for this tool.

fn prepare<'life0, 'life1, 'async_trait>( &'life0 self, _ctx: &'life1 RunContext<Deps>, def: ToolDefinition, ) -> Pin<Box<dyn Future<Output = Option<ToolDefinition>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Prepare the tool definition at runtime.

fn name(&self) -> String

Get the tool name.

fn description(&self) -> String

Get the tool description.