Hare/yoi
1
1
Fork 0
Code Actions Packages Releases Activity

fix: bound task tool output

Browse Source
This commit is contained in:
Keisuke Hirata 2026-07-29 02:29:15 +09:00
parent 450e0cddbd
commit b6f83d81bb
No known key found for this signature in database
5 changed files with 240 additions and 89 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

66
crates/tui/src/task.rs
View File

@ -27,6 +27,12 @@ pub enum TaskStatus {
Deleted,
}
impl TaskStatus {
fn is_active(self) -> bool {
matches!(self, Self::Pending | Self::Inprogress)
}
}
#[derive(Debug, Clone, PartialEq, Eq, Deserialize)]
pub struct TaskEntry {
pub taskid: u64,
@ -106,8 +112,10 @@ impl TaskStore {
}
"TaskUpdate" => {
if let Ok(p) = serde_json::from_str::<TaskUpdateParams>(arguments)
&& let Some(t) = self.tasks.iter_mut().find(|t| t.taskid == p.taskid)
&& let Some(task_position) =
self.tasks.iter().position(|t| t.taskid == p.taskid)
{
let t = &mut self.tasks[task_position];
if let Some(s) = p.status {
t.status = s;
}
@ -117,6 +125,9 @@ impl TaskStore {
if let Some(d) = p.description {
t.description = d;
}
if !t.status.is_active() {
self.tasks.remove(task_position);
}
}
}
_ => {}
@ -132,6 +143,10 @@ impl TaskStore {
}
fn replace_with(&mut self, tasks: Vec<TaskEntry>) {
let tasks: Vec<_> = tasks
.into_iter()
.filter(|task| task.status.is_active())
.collect();
self.next_taskid = tasks
.iter()
.map(|t| t.taskid)
@ -175,7 +190,13 @@ fn parse_snapshot_text(text: &str) -> Option<Vec<TaskEntry>> {
let rest = &text[start..];
let end = rest.find(end_marker)?;
let snapshot: TaskSnapshot = serde_json::from_str(&rest[..end]).ok()?;
Some(snapshot.tasks)
Some(
snapshot
.tasks
.into_iter()
.filter(|task| task.status.is_active())
.collect(),
)
}
#[cfg(test)]
@ -220,7 +241,7 @@ mod tests {
}
#[test]
fn counts_classifies_each_status() {
fn counts_tracks_only_active_tasks_after_completion() {
let mut s = TaskStore::new();
s.apply_tool_call("TaskCreate", r#"{"subject":"a","description":""}"#);
s.apply_tool_call("TaskCreate", r#"{"subject":"b","description":""}"#);
@ -230,9 +251,9 @@ mod tests {
let c = s.counts();
assert_eq!(c.pending, 1);
assert_eq!(c.inprogress, 1);
assert_eq!(c.completed, 1);
assert_eq!(c.completed, 0);
assert_eq!(c.deleted, 0);
assert_eq!(c.total(), 3);
assert_eq!(c.total(), 2);
assert_eq!(c.active(), 2);
}
@ -242,7 +263,7 @@ mod tests {
fn wrap_snapshot(json_body: &str, overview: &str) -> String {
format!(
"[Session TaskStore snapshot]\n\n{overview}\n\n```json\n{json_body}\n```\n\n\
This is the complete session task list preserved across compaction. \
This is the active session task list preserved across compaction. \
The following TaskList tool result presents the same state through the tool lane."
)
}
@ -267,20 +288,19 @@ mod tests {
}"#;
let text = wrap_snapshot(
body,
"TaskStore: 2 task(s) (pending: 1, inprogress: 0, completed: 1, deleted: 0)",
"TaskStore: 1 active task(s) (pending: 1, inprogress: 0)",
);
let mut s = TaskStore::new();
s.apply_tool_call("TaskCreate", r#"{"subject":"stale","description":""}"#);
s.apply_system_message_text(&text);
let tasks = s.tasks();
assert_eq!(tasks.len(), 2);
assert_eq!(tasks[0].taskid, 5);
assert_eq!(tasks[0].status, TaskStatus::Completed);
assert_eq!(tasks[1].taskid, 7);
// Subsequent TaskCreate must continue beyond the highest taskid
assert_eq!(tasks.len(), 1);
assert_eq!(tasks[0].taskid, 7);
assert_eq!(tasks[0].status, TaskStatus::Pending);
// Subsequent TaskCreate must continue beyond the highest active taskid
// observed in the snapshot.
s.apply_tool_call("TaskCreate", r#"{"subject":"new","description":""}"#);
assert_eq!(s.tasks()[2].taskid, 8);
assert_eq!(s.tasks()[1].taskid, 8);
}
#[test]
@ -304,7 +324,7 @@ mod tests {
}
]
}"#;
let text = wrap_snapshot(body, "TaskStore: 1 task(s)");
let text = wrap_snapshot(body, "TaskStore: 1 active task(s)");
let mut s = TaskStore::new();
s.apply_system_message_text(&text);
let t = &s.tasks()[0];
@ -329,13 +349,13 @@ mod snapshot_format_contract {
fn wrap_pod_style(snapshot_text: &str) -> String {
format!(
"[Session TaskStore snapshot]\n\n{snapshot_text}\n\n\
This is the complete session task list preserved across compaction. \
This is the active session task list preserved across compaction. \
The following TaskList tool result presents the same state through the tool lane."
)
}
fn snapshot_fixture() -> &'static str {
r#"TaskStore: 2 task(s) (pending: 0, inprogress: 1, completed: 1, deleted: 0)
r#"TaskStore: 1 active task(s) (pending: 0, inprogress: 1)
```json
{
@ -345,12 +365,6 @@ mod snapshot_format_contract {
"status": "inprogress",
"subject": "first",
"description": "first desc"
},
{
"taskid": 2,
"status": "completed",
"subject": "second",
"description": "second desc with\nnewline"
}
]
}
@ -358,7 +372,7 @@ mod snapshot_format_contract {
}
fn empty_snapshot_fixture() -> &'static str {
r#"TaskStore: 0 task(s) (pending: 0, inprogress: 0, completed: 0, deleted: 0)
r#"TaskStore: 0 active task(s) (pending: 0, inprogress: 0)
```json
{
@ -384,15 +398,11 @@ mod snapshot_format_contract {
downstream.apply_system_message_text(&envelope);
let tasks = downstream.tasks();
assert_eq!(tasks.len(), 2, "TUI parsed wrong number of tasks");
assert_eq!(tasks.len(), 1, "TUI parsed wrong number of tasks");
assert_eq!(tasks[0].taskid, 1);
assert_eq!(tasks[0].subject, "first");
assert_eq!(tasks[0].description, "first desc");
assert_eq!(status_label(tasks[0].status), "inprogress");
assert_eq!(tasks[1].taskid, 2);
assert_eq!(tasks[1].subject, "second");
assert_eq!(tasks[1].description, "second desc with\nnewline");
assert_eq!(status_label(tasks[1].status), "completed");
}
#[test]

4
crates/worker/src/feature/builtin/task/mod.rs
View File

@ -76,7 +76,7 @@ impl TaskFeature {
/// pointing at the same feature-owned store after rewind.
pub fn restore_from_history(&self, history: &[Item]) {
let restored = TaskStore::from_history(history);
self.state.task_store.replace_with(restored.list());
self.state.task_store.replace_with(restored.list_active());
}
/// Feature-owned snapshot text used by compaction to preserve Task state.
@ -86,7 +86,7 @@ impl TaskFeature {
/// Feature-owned compact summary used for the synthetic TaskList result.
pub fn snapshot_overview(&self) -> String {
snapshot_overview(&self.state.task_store.list())
snapshot_overview(&self.state.task_store.list_active())
}
#[cfg(test)]

111
crates/worker/src/feature/builtin/task/store.rs
View File

@ -18,6 +18,12 @@ pub enum TaskStatus {
Deleted,
}
impl TaskStatus {
pub fn is_active(self) -> bool {
matches!(self, Self::Pending | Self::Inprogress)
}
}
impl std::fmt::Display for TaskStatus {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
let s = match self {
@ -54,6 +60,8 @@ pub struct TaskSnapshot {
pub tasks: Vec<TaskEntry>,
}
pub const DEFAULT_TASK_LIST_LIMIT: usize = 20;
impl TaskStore {
pub fn new() -> Self {
Self {
@ -85,6 +93,13 @@ impl TaskStore {
.clone()
}
pub fn list_active(&self) -> Vec<TaskEntry> {
self.list()
.into_iter()
.filter(|task| task.status.is_active())
.collect()
}
pub fn get(&self, taskid: u64) -> Option<TaskEntry> {
self.inner
.lock()
@ -106,11 +121,12 @@ impl TaskStore {
return Err(TaskStoreError::NoFields);
}
let mut inner = self.inner.lock().unwrap_or_else(|e| e.into_inner());
let task = inner
let task_position = inner
.tasks
.iter_mut()
.find(|t| t.taskid == taskid)
.iter()
.position(|t| t.taskid == taskid)
.ok_or(TaskStoreError::Missing(taskid))?;
let task = &mut inner.tasks[task_position];
if let Some(status) = status {
task.status = status;
}
@ -120,11 +136,21 @@ impl TaskStore {
if let Some(description) = description {
task.description = description;
}
Ok(task.clone())
let updated = task.clone();
if !updated.status.is_active() {
inner.tasks.remove(task_position);
}
Ok(updated)
}
pub fn snapshot(&self) -> TaskSnapshot {
TaskSnapshot { tasks: self.list() }
self.snapshot_limited(DEFAULT_TASK_LIST_LIMIT)
}
pub fn snapshot_limited(&self, limit: usize) -> TaskSnapshot {
TaskSnapshot {
tasks: self.list_active().into_iter().take(limit).collect(),
}
}
pub fn replay_history(&self, history: &[Item]) {
@ -168,6 +194,10 @@ impl TaskStore {
}
pub fn replace_with(&self, tasks: Vec<TaskEntry>) {
let tasks: Vec<_> = tasks
.into_iter()
.filter(|task| task.status.is_active())
.collect();
let next_taskid = tasks
.iter()
.map(|t| t.taskid)
@ -237,27 +267,26 @@ pub fn snapshot_overview(tasks: &[TaskEntry]) -> String {
.iter()
.filter(|t| t.status == TaskStatus::Inprogress)
.count();
let completed = tasks
.iter()
.filter(|t| t.status == TaskStatus::Completed)
.count();
let deleted = tasks
.iter()
.filter(|t| t.status == TaskStatus::Deleted)
.count();
format!(
"TaskStore: {} task(s) (pending: {pending}, inprogress: {inprogress}, completed: {completed}, deleted: {deleted})",
tasks.len()
)
let active = pending + inprogress;
format!("TaskStore: {active} active task(s) (pending: {pending}, inprogress: {inprogress})")
}
pub fn render_snapshot(tasks: &[TaskEntry]) -> String {
let active_tasks: Vec<_> = tasks
.iter()
.filter(|task| task.status.is_active())
.cloned()
.collect();
let snapshot = TaskSnapshot {
tasks: tasks.to_vec(),
tasks: active_tasks.clone(),
};
let json =
serde_json::to_string_pretty(&snapshot).unwrap_or_else(|_| String::from("{\"tasks\":[]}"));
format!("{}\n\n```json\n{}\n```\n", snapshot_overview(tasks), json)
format!(
"{}\n\n```json\n{}\n```\n",
snapshot_overview(&active_tasks),
json
)
}
pub(super) fn parse_compact_snapshot_text(text: &str) -> Option<Vec<TaskEntry>> {
@ -270,7 +299,13 @@ pub(super) fn parse_compact_snapshot_text(text: &str) -> Option<Vec<TaskEntry>>
let rest = &text[start..];
let end = rest.find(end_marker)?;
let snapshot: TaskSnapshot = serde_json::from_str(&rest[..end]).ok()?;
Some(snapshot.tasks)
Some(
snapshot
.tasks
.into_iter()
.filter(|task| task.status.is_active())
.collect(),
)
}
#[cfg(test)]
@ -288,11 +323,9 @@ mod tests {
];
let store = TaskStore::from_history(&history);
let tasks = store.list();
assert_eq!(tasks.len(), 2);
assert_eq!(tasks.len(), 1);
assert_eq!(tasks[0].taskid, 1);
assert_eq!(tasks[0].status, TaskStatus::Pending);
assert_eq!(tasks[1].taskid, 2);
assert_eq!(tasks[1].status, TaskStatus::Completed);
}
/// Wrap snapshot text the way `Worker::try_pre_run_compact` does, so tests
@ -300,7 +333,7 @@ mod tests {
fn wrap_snapshot_system_message(snapshot: &str) -> String {
format!(
"[Session TaskStore snapshot]\n\n{snapshot}\n\n\
This is the complete session task list preserved across compaction. \
This is the active session task list preserved across compaction. \
The following TaskList tool result presents the same state through the tool lane."
)
}
@ -322,11 +355,9 @@ mod tests {
];
let store = TaskStore::from_history(&history);
let tasks = store.list();
assert_eq!(tasks.len(), 2);
assert_eq!(tasks[0].taskid, 1);
assert_eq!(tasks[0].status, TaskStatus::Completed);
assert_eq!(tasks[1].taskid, 2);
assert_eq!(tasks[1].subject, "new");
assert_eq!(tasks.len(), 1);
assert_eq!(tasks[0].taskid, 2);
assert_eq!(tasks[0].subject, "new");
}
#[test]
@ -365,15 +396,12 @@ mod tests {
];
let store = TaskStore::from_history(&history);
let tasks = store.list();
assert_eq!(tasks.len(), 3);
assert_eq!(tasks[0].taskid, 1);
assert_eq!(tasks[0].subject, "A");
assert_eq!(tasks[0].status, TaskStatus::Completed);
assert_eq!(tasks[1].taskid, 2);
assert_eq!(tasks[1].subject, "B");
assert_eq!(tasks[1].status, TaskStatus::Inprogress);
assert_eq!(tasks[2].taskid, 3);
assert_eq!(tasks[2].subject, "C");
assert_eq!(tasks.len(), 2);
assert_eq!(tasks[0].taskid, 2);
assert_eq!(tasks[0].subject, "B");
assert_eq!(tasks[0].status, TaskStatus::Inprogress);
assert_eq!(tasks[1].taskid, 3);
assert_eq!(tasks[1].subject, "C");
}
#[test]
@ -415,9 +443,10 @@ mod tests {
let snapshot_text = pre.snapshot_text();
let system = Item::system_message(wrap_snapshot_system_message(&snapshot_text));
let call = Item::tool_call("compact-tasklist", "TaskList", "{}");
let active_tasks = pre.list_active();
let result = Item::tool_result_with_content(
"compact-tasklist",
snapshot_overview(&pre.list()),
snapshot_overview(&active_tasks),
snapshot_text.clone(),
);
@ -426,7 +455,7 @@ mod tests {
.as_text()
.and_then(parse_compact_snapshot_text)
.expect("system message should parse as snapshot");
assert_eq!(extracted, pre.list());
assert_eq!(extracted, active_tasks);
// The synthetic call/result pair shares one call_id and carries the
// expected tool name + detailed content.
@ -453,6 +482,6 @@ mod tests {
// Replaying the full triple reconstructs the same TaskStore.
let store = TaskStore::from_history(&[system, call, result]);
assert_eq!(store.list(), pre.list());
assert_eq!(store.list(), active_tasks);
}
}

146
crates/worker/src/feature/builtin/task/tool_impl.rs
View File

@ -6,7 +6,7 @@ use async_trait::async_trait;
use llm_engine::tool::{Tool, ToolDefinition, ToolError, ToolMeta, ToolOutput};
use serde::Deserialize;
use super::store::{TaskEntry, TaskStatus, TaskStore, render_snapshot, snapshot_overview};
use super::store::{DEFAULT_TASK_LIST_LIMIT, TaskEntry, TaskStatus, TaskStore, snapshot_overview};
#[derive(Debug, Deserialize, schemars::JsonSchema)]
struct TaskCreateParams {
@ -17,7 +17,11 @@ struct TaskCreateParams {
}
#[derive(Debug, Deserialize, schemars::JsonSchema)]
struct TaskListParams {}
struct TaskListParams {
/// Maximum number of active tasks to return. Defaults to 20.
#[serde(default)]
limit: Option<usize>,
}
#[derive(Debug, Deserialize, schemars::JsonSchema)]
struct TaskGetParams {
@ -58,9 +62,7 @@ coordination. Do not create a task just because a request has several steps, and
one for short questions, quick checks, single reviews, or one-off commands. Prefer updating an \
existing active task over creating a duplicate. Input only `subject` and `description`; `taskid` \
is assigned automatically and initial `status` is `pending`.";
const LIST_DESCRIPTION: &str = "List every session-lifetime task, including completed and \
deleted entries. Tasks are user-visible real-time status for short-term current-work tracking. \
Takes an empty object as input.";
const LIST_DESCRIPTION: &str = "List active session-lifetime tasks. Completed and deleted tasks are forgotten and omitted. Defaults to 20 tasks unless `limit` is provided.";
const GET_DESCRIPTION: &str = "Get one session-lifetime task by `taskid`. Tasks are \
user-visible real-time status for short-term current-work tracking. Returns an error if the task \
does not exist.";
@ -81,7 +83,7 @@ impl Tool for TaskCreateTool {
let params: TaskCreateParams = serde_json::from_str(input_json)
.map_err(|e| ToolError::InvalidArgument(format!("invalid TaskCreate input: {e}")))?;
let created = self.store.create(params.subject, params.description);
let tasks = self.store.list();
let tasks = self.store.list_active();
Ok(task_output(
format!(
"Created task {} ({})\n{}",
@ -90,7 +92,6 @@ impl Tool for TaskCreateTool {
snapshot_overview(&tasks)
),
&created,
&tasks,
))
}
}
@ -102,12 +103,14 @@ impl Tool for TaskListTool {
input_json: &str,
_ctx: llm_engine::tool::ToolExecutionContext,
) -> Result<ToolOutput, ToolError> {
let _: TaskListParams = serde_json::from_str(input_json)
let params: TaskListParams = serde_json::from_str(input_json)
.map_err(|e| ToolError::InvalidArgument(format!("invalid TaskList input: {e}")))?;
let tasks = self.store.list();
let limit = params.limit.unwrap_or(DEFAULT_TASK_LIST_LIMIT);
let active_tasks = self.store.list_active();
let tasks: Vec<_> = active_tasks.iter().take(limit).cloned().collect();
Ok(ToolOutput {
summary: snapshot_overview(&tasks),
content: Some(render_snapshot(&tasks)),
summary: list_overview(active_tasks.len(), tasks.len()),
content: Some(render_task_list(&tasks)),
})
}
}
@ -150,7 +153,7 @@ impl Tool for TaskUpdateTool {
params.description,
)
.map_err(|e| ToolError::ExecutionFailed(e.to_string()))?;
let tasks = self.store.list();
let tasks = self.store.list_active();
Ok(task_output(
format!(
"Updated task {} ({})\n{}",
@ -159,21 +162,35 @@ impl Tool for TaskUpdateTool {
snapshot_overview(&tasks)
),
&updated,
&tasks,
))
}
}
fn task_output(summary: String, task: &TaskEntry, tasks: &[TaskEntry]) -> ToolOutput {
fn task_output(summary: String, task: &TaskEntry) -> ToolOutput {
let content = serde_json::json!({
"task": task,
"snapshot": { "tasks": tasks },
});
ToolOutput {
summary,
content: Some(serde_json::to_string_pretty(&content).unwrap_or_default()),
}
}
fn list_overview(total_active: usize, returned: usize) -> String {
if returned < total_active {
format!(
"TaskStore: {returned} active task(s) shown; {} omitted.",
total_active - returned
)
} else {
format!("TaskStore: {returned} active task(s)")
}
}
fn render_task_list(tasks: &[TaskEntry]) -> String {
serde_json::to_string_pretty(tasks).unwrap_or_default()
}
fn task_create_tool(store: TaskStore) -> ToolDefinition {
Arc::new(move || {
let schema = schemars::schema_for!(TaskCreateParams);
@ -286,10 +303,105 @@ mod tests {
assert!(out.content.unwrap().contains("implement tasks"));
let out = list.execute("{}", Default::default()).await.unwrap();
assert!(out.summary.contains("1 task(s)"));
assert!(out.summary.contains("1 active task(s)"));
let content = out.content.unwrap();
assert!(content.contains("\"taskid\": 1"));
assert!(content.contains("```json"));
assert!(!content.contains("\"limit\""));
assert!(!content.contains("```json"));
}
#[tokio::test]
async fn task_list_omits_completed_deleted_and_defaults_to_twenty() {
let store = TaskStore::new();
let create = tool(task_create_tool(store.clone()));
let update = tool(task_update_tool(store.clone()));
let list = tool(task_list_tool(store.clone()));
for i in 0..25 {
create
.execute(
&format!(r#"{{"subject":"task {i}","description":"desc {i}"}}"#),
Default::default(),
)
.await
.unwrap();
}
update
.execute(r#"{"taskid":1,"status":"completed"}"#, Default::default())
.await
.unwrap();
update
.execute(r#"{"taskid":2,"status":"deleted"}"#, Default::default())
.await
.unwrap();
let out = list.execute("{}", Default::default()).await.unwrap();
assert_eq!(
out.summary,
"TaskStore: 20 active task(s) shown; 3 omitted."
);
let content = out.content.unwrap();
let json: serde_json::Value = serde_json::from_str(&content).unwrap();
let tasks = json.as_array().unwrap();
assert_eq!(tasks.len(), 20);
let ids: Vec<u64> = tasks
.iter()
.map(|task| task["taskid"].as_u64().unwrap())
.collect();
assert!(!ids.contains(&1));
assert!(!ids.contains(&2));
assert!(!content.contains("\"limit\""));
assert!(!content.contains("\"total_active\""));
assert!(!content.contains("\"truncated\""));
let out = list
.execute(r#"{"limit":3}"#, Default::default())
.await
.unwrap();
assert_eq!(
out.summary,
"TaskStore: 3 active task(s) shown; 20 omitted."
);
let content = out.content.unwrap();
let json: serde_json::Value = serde_json::from_str(&content).unwrap();
assert_eq!(json.as_array().unwrap().len(), 3);
}
#[tokio::test]
async fn task_create_and_update_results_do_not_include_full_snapshot() {
let store = TaskStore::new();
let create = tool(task_create_tool(store.clone()));
let update = tool(task_update_tool(store.clone()));
create
.execute(
r#"{"subject":"done","description":"completed task"}"#,
Default::default(),
)
.await
.unwrap();
update
.execute(r#"{"taskid":1,"status":"completed"}"#, Default::default())
.await
.unwrap();
create
.execute(
r#"{"subject":"active","description":"active task"}"#,
Default::default(),
)
.await
.unwrap();
let out = update
.execute(r#"{"taskid":2,"status":"inprogress"}"#, Default::default())
.await
.unwrap();
let content = out.content.unwrap();
assert!(content.contains("\"task\""));
assert!(content.contains("\"taskid\": 2"));
assert!(!content.contains("\"snapshot\""));
assert!(!content.contains("\"taskid\": 1"));
assert!(!content.contains("completed task"));
}
#[tokio::test]

2
crates/worker/src/worker.rs
View File

@ -2733,7 +2733,7 @@ impl<C: LlmClient, St: Store> Worker<C, St> {
}
let task_snapshot_message = Item::system_message(format!(
"[Session TaskStore snapshot]\n\n{task_snapshot_text}\n\n\
This is the complete session task list preserved across compaction. \
This is the active session task list preserved across compaction. \
The following TaskList tool result presents the same state through the tool lane."
));
compact_introduced_system_messages.push(task_snapshot_message.clone());