Enum AgentAction 

pub enum AgentAction {
    ReadFile {
        paths: Vec<String>,
    },
    WriteFile {
        path: String,
        content: String,
    },
    EditFile {
        path: String,
        old_string: String,
        new_string: String,
    },
    DeleteFile {
        path: String,
    },
    CreateDirectory {
        path: String,
    },
    ExecuteCommand {
        command: String,
        working_dir: Option<String>,
        timeout: Option<u64>,
    },
    WebSearch {
        queries: Vec<(String, usize)>,
    },
    WebFetch {
        url: String,
    },
    SpawnAgent {
        prompt: String,
        description: String,
    },
    Screenshot {
        mode: String,
        monitor: Option<String>,
        region: Option<String>,
        window: Option<String>,
    },
    Click {
        x: i32,
        y: i32,
        button: String,
    },
    TypeText {
        text: String,
    },
    PressKey {
        key: String,
    },
    Scroll {
        direction: String,
        amount: i32,
    },
    MouseMove {
        x: i32,
        y: i32,
    },
    ListWindows,
    McpToolCall {
        server_name: String,
        tool_name: String,
        arguments: Value,
    },
    ParseError {
        message: String,
    },
}

Represents an action that the AI wants to perform

Variants

ReadFile

Read one or more files (executor decides parallelization)

Fields

paths: Vec<String>

WriteFile

Write or create a file

Fields

path: String

content: String

EditFile

Make targeted edits to a file by replacing specific text

Fields

path: String

old_string: String

new_string: String

DeleteFile

Delete a file

Fields

path: String

CreateDirectory

Create a directory

Fields

path: String

ExecuteCommand

Execute a shell command

Fields

command: String

working_dir: Option<String>

timeout: Option<u64>

WebSearch

Web search via Ollama Cloud API (executor decides parallelization)

Fields

queries: Vec<(String, usize)>

WebFetch

Fetch a URL’s content via Ollama Cloud API

Fields

url: String

SpawnAgent

Spawn an autonomous sub-agent with its own conversation context

Fields

prompt: String

description: String

Screenshot

Capture a screenshot of the screen (or a focused window/monitor/region/window)

Fields

mode: String

monitor: Option<String>

region: Option<String>

window: Option<String>

Click

Click at screen coordinates

Fields

x: i32

y: i32

button: String

TypeText

Type a text string at the current cursor position

Fields

text: String

PressKey

Press a key or key combination

Fields

key: String

Scroll

Scroll in a direction

Fields

direction: String

amount: i32

MouseMove

Move mouse cursor to coordinates

Fields

x: i32

y: i32

ListWindows

List all visible window titles (lightweight, no screenshot)

McpToolCall

Dynamic MCP tool call (dispatched to an MCP server at runtime)

Fields

server_name: String

tool_name: String

arguments: Value

ParseError

Placeholder for tool calls that failed to parse (never executed)

Fields

message: String

Implementations

impl AgentAction

pub fn display_info(&self) -> (&str, String)

Extract a (type_label, target) pair for display or logging

Trait Implementations

impl Clone for AgentAction

fn clone(&self) -> AgentAction

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for AgentAction

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for AgentAction

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for AgentAction

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.