jobq: complete is the completion, and failure grows nothing

Two review findings from the previous round, re-checked against the
actual tree rather than against my notes.

`Scheduler::complete` was still a public wrapper whose entire body was
`self.finish(id, outcome)`. Its docstring argued the split was not a
redirect because both completion forms shared `finish` -- but sharing a
private helper is not a reason for two public names. `finish`'s body now
lives in `complete`, and `complete_growing` calls it. Same sharing, one
name, no redirect.

Growth on a failed node is now dropped by `complete_growing` instead of
by the host loop. Failure cancel-cascades to every pending child of the
completing node, and grown work is inserted as its children, so anything
appended here is Skipped by the next statement -- the insert is not
wrong, it is provably pointless. That is a consequence of this crate's
cascade rule, so this crate should be the one enforcing it; a host that
has to remember it can forget it. Behaviour is unchanged: hive-c0re
already dropped growth before calling, and now no longer has to.

`Scheduler::new_job` is left alone but documented for what it is: the
hole in `JobBuilder::new`'s pub(crate) wall, with no non-test caller
since claim_next mints a builder per running node. Closing it is a venue
question rather than a rename, so it stays for now.

hive-c0re/src/job_queue/mod.rs

@@ -252,22 +252,6 @@ impl JobQueue {
         self.lock().graph().root_of(node).map(NodeId::get)
     }
 
-    /// A builder for a node to declare more work into while it runs.
-    ///
-    /// Handed to [`exec::run_node`] and returned to the crate's completion.
-    /// Only `hive_jobq` can construct one, which is why this goes through the
-    /// scheduler rather than `Job::default()`.
-    ///
-    /// The completion wrappers that used to live beside this (`complete_node`,
-    /// `complete_node_growing`) are **gone**: a node is completed inside the
-    /// future [`hive_jobq::scheduler::Scheduler::claim_next`] hands back, so
-    /// this layer has nothing left to wrap. The tests keep their own extension
-    /// trait for driving completions by hand.
-    #[must_use]
-    pub fn new_job(&self) -> Job {
-        self.lock().new_job()
-    }
-
     /// Cancel a DAG that hasn't started yet: every work node is still `Pending`,
     /// so each is cancelled. `false` once any work node is running or terminal —
     /// an in-flight nix build isn't interruptible.

hive-c0re/src/job_queue/scheduler.rs

@@ -101,17 +101,14 @@ pub async fn run_worker(coord: Arc<Coordinator>) {
                             "job_queue: node failed"
                         ),
                     }
-                    let outcome = super::outcome_of(result.map_err(|e| format!("{e:#}")));
-                    // Growth is dropped on failure: a node that declared
-                    // follow-up work and *then* failed does not want it run —
-                    // failure cancel-cascades, so inserting it would only add
-                    // nodes to immediately cancel.
-                    let grown = if matches!(outcome, hive_jobq::scheduler::Outcome::Failed(_)) {
-                        coord.job_queue.new_job()
-                    } else {
-                        grown
-                    };
-                    (grown, outcome)
+                    // Growth on a failed node is dropped by `complete_growing`,
+                    // not here: failure cancel-cascades inside jobq, so that
+                    // rule is the crate's to enforce and this loop does not get
+                    // to forget it.
+                    (
+                        grown,
+                        super::outcome_of(result.map_err(|e| format!("{e:#}"))),
+                    )
                 }
             })
         };

hive-c0re/src/job_queue/tests.rs

@@ -122,9 +122,23 @@ impl ClaimReady for JobQueue {
 trait CompleteNode {
     fn complete_node(&self, node_id: NodeId, result: Result<(), String>);
     fn complete_node_growing(&self, node_id: NodeId, result: Result<(), String>, grown: Job);
+    fn new_job(&self) -> Job;
 }
 
 impl CompleteNode for JobQueue {
+    /// Mint a builder to declare growth into.
+    ///
+    /// Test-only for the same reason as the rest of this trait: production
+    /// never mints one, because `claim_next` hands each running node its
+    /// builder and takes it back. That leaves `Scheduler::new_job` with no
+    /// non-test caller either — see the note on that fn.
+    fn new_job(&self) -> Job {
+        self.sched()
+            .lock()
+            .expect("job_queue mutex poisoned")
+            .new_job()
+    }
+
     fn complete_node(&self, node_id: NodeId, result: Result<(), String>) {
         self.sched()
             .lock()