docs: clarify name pool vs polecat pool misconception

Fix misleading language that could suggest polecats wait in an idle pool:

- refinery/engineer.go: "available polecat" → "fresh polecat (spawned on demand)"
- namepool.go: Clarify this pools NAMES not polecats; polecats are spawned
  fresh and nuked when done, only name slots are reused
- dog-pool-architecture.md: "Pool allocation pattern" → "Name slot allocation
  pattern (pool of names, not instances)"

There is no idle pool of polecats. They are spawned for work and nuked when done.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

dog-pool-architecture.md

@@ -306,7 +306,7 @@ Rationale:
 What dogs DO share:
 - tmux utilities for message sending/capture
 - State file patterns
-- Pool allocation pattern
+- Name slot allocation pattern (pool of names, not instances)
 
 ### Dog Execution Loop
 

internal/polecat/namepool.go

@@ -12,7 +12,10 @@ import (
 )
 
 const (
-	// DefaultPoolSize is the number of reusable names in the pool.
+	// DefaultPoolSize is the number of name slots in the pool.
+	// NOTE: This is a pool of NAMES, not polecats. Polecats are spawned fresh
+	// for each task and nuked when done - there is no idle pool of polecats.
+	// Only the name slots are reused when a polecat is nuked and a new one spawned.
 	DefaultPoolSize = 50
 
 	// DefaultTheme is the default theme for new rigs.
@@ -59,7 +62,12 @@ var BuiltinThemes = map[string][]string{
 	},
 }
 
-// NamePool manages a bounded pool of reusable polecat names.
+// NamePool manages a bounded pool of reusable polecat NAME SLOTS.
+// IMPORTANT: This pools NAMES, not polecats. Polecats are spawned fresh for each
+// task and nuked when done - there is no idle pool of polecat instances waiting
+// for work. When a polecat is nuked, its name slot becomes available for the next
+// freshly-spawned polecat.
+//
 // Names are drawn from a themed pool (mad-max by default).
 // When the pool is exhausted, overflow names use rigname-N format.
 type NamePool struct {
@@ -214,7 +222,9 @@ func (p *NamePool) Allocate() (string, error) {
 	return name, nil
 }
 
-// Release returns a pooled name to the pool.
+// Release returns a name slot to the available pool.
+// Called when a polecat is nuked - the name becomes available for new polecats.
+// NOTE: This releases the NAME, not the polecat. The polecat is gone (nuked).
 // For overflow names, this is a no-op (they are not reusable).
 func (p *NamePool) Release(name string) {
 	p.mu.Lock()

internal/refinery/engineer.go

@@ -632,7 +632,7 @@ func (e *Engineer) handleFailureFromQueue(mr *mrqueue.MR, result ProcessResult)
 }
 
 // createConflictResolutionTask creates a dispatchable task for resolving merge conflicts.
-// This task will be picked up by bd ready and can be dispatched to an available polecat.
+// This task will be picked up by bd ready and can be slung to a fresh polecat (spawned on demand).
 // Returns the created task's ID for blocking the MR until resolution.
 //
 // Task format: