Struct TaskLoopContext 

pub struct TaskLoopContext {
    pub session_id: String,
    pub items: Vec<TaskLoopItem>,
    pub active_item_id: Option<String>,
    pub current_round: u32,
    pub max_rounds: u32,
    pub created_at: DateTime<Utc>,
    pub updated_at: DateTime<Utc>,
    pub version: u64,
    pub task_list_dirty: bool,
}

TaskList context for Agent Loop

Acts as a first-class citizen in the agent loop, tracking task progress throughout the entire conversation lifecycle.

Fields

session_id: String

Session ID

items: Vec<TaskLoopItem>

Task items with execution tracking

active_item_id: Option<String>

Currently active task item ID

current_round: u32

Current round number

max_rounds: u32

Maximum rounds allowed

created_at: DateTime<Utc>

Creation timestamp

updated_at: DateTime<Utc>

Last update timestamp

version: u64

Version number for conflict detection

task_list_dirty: bool

Set when the Task tool structurally rewrites the list, cleared once an evaluation has been spawned for that change. This is the single signal that gates async task evaluation to actual Task-tool writes, instead of firing every round of tool activity (which bumps version without changing the plan).

Implementations

impl TaskLoopContext

pub fn auto_match_tool_to_item(&mut self, tool_name: &str)

Auto-match tool to task item based on keywords

pub fn auto_update_status(&mut self, tool_name: &str, result: &ToolResult)

Auto-update status based on tool execution result

impl TaskLoopContext

pub fn from_session(session: &Session) -> Option<Self>

Create TaskLoopContext from the session’s task list.

pub fn to_task_list_with_title(&self, title: impl Into<String>) -> TaskList

Clone into a TaskList, preserving the provided title.

pub fn into_task_list(self) -> TaskList

Convert back to TaskList for persistence.

pub fn into_task_list_with_title(self, title: impl Into<String>) -> TaskList

Convert back to TaskList for persistence with an explicit title.

impl TaskLoopContext

pub fn format_for_prompt(&self) -> String

Generate context for prompt injection

impl TaskLoopContext

pub fn track_tool_execution( &mut self, tool_name: &str, result: &ToolResult, round: u32, )

Track tool execution

Records a tool call and associates it with the active task item.

pub fn set_active_item(&mut self, item_id: &str)

Set active task item

Moves the previous active item back to pending and activates a new item.

pub fn update_item_status(&mut self, item_id: &str, status: TaskItemStatus)

Update item status manually

pub fn apply_evaluated_update( &mut self, item_id: &str, status: TaskItemStatus, notes: Option<&str>, evidence: Option<&str>, blocker: Option<&str>, ) -> bool

Apply an evaluated task update and report whether it changed the context.

pub fn append_structured_feedback( &mut self, item_id: &str, evidence: Option<&str>, blocker: Option<&str>, )

impl TaskLoopContext

pub fn is_all_completed(&self) -> bool

Check if all items are completed

Trait Implementations

impl Clone for TaskLoopContext

fn clone(&self) -> TaskLoopContext

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TaskLoopContext

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

Formats the value using the given formatter.