refactor(#2569): remove the c0re todo store + handlers (todos now owned in-container)

hive-c0re/src/loose_ends.rs

@@ -97,40 +97,9 @@ pub fn for_agent(coord: &Coordinator, agent: &str) -> Result<Vec<LooseEnd>> {
             age_seconds: saturating_age(now, r.created_at.timestamp()),
         });
     }
-    // Dynamic, subsystem-pushed todos (loose-ends v2). Scoped to
-    // this agent; the producing subsystem or the agent itself clears them.
-    out.extend(todos_for(coord, agent, None)?);
     Ok(out)
 }
 
-/// This agent's dynamic todos as `LooseEnd::Todo` rows, optionally
-/// filtered to one `subsystem`. Shared by [`for_agent`] and the
-/// `ListTodos` handler so the row-mapping lives in one place.
-///
-/// # Errors
-///
-/// Propagates the todo-store query failure.
-pub fn todos_for(
-    coord: &Coordinator,
-    agent: &str,
-    subsystem: Option<&str>,
-) -> Result<Vec<LooseEnd>> {
-    let now = now_unix();
-    Ok(coord
-        .todos
-        .list(agent, subsystem)?
-        .into_iter()
-        .map(|t| LooseEnd::Todo {
-            id: t.id,
-            subsystem: t.subsystem,
-            subsystem_key: t.subsystem_key,
-            summary: t.summary,
-            source: t.source,
-            age_seconds: saturating_age(now, t.updated_at.timestamp()),
-        })
-        .collect())
-}
-
 /// Hive-wide loose-ends view: EVERY pending approval + EVERY
 /// unanswered question + EVERY pending reminder. Manager surface
 /// only; sub-agents can't see each other's threads via the agent