fix: flat-budget walk in get_memories_for_context

The per-type slicing (available // len(memory_types)) starved
paragraph-length memories: with 3 types and a 450-char budget,
each type got ~131 chars while real project memories are 300-500
chars each — every entry was skipped and the new Project Memories
band never appeared in the live pack.

Switch to a flat budget pool walked type-by-type in order. Short
identity/preference memories still get first pick when the budget
is tight, but long project memories can now compete for space.

Caught on the first post-deploy probe: 2 active p04 memories
existed but none landed in formatted_context.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

src/atocore/memory/service.py

@@ -369,10 +369,13 @@ def get_memories_for_context(
 
     available = budget - wrapper_chars
     selected_entries: list[str] = []
+    used = 0
 
-    for index, mtype in enumerate(memory_types):
-        type_budget = available if index == len(memory_types) - 1 else max(0, available // (len(memory_types) - index))
-        type_used = 0
+    # Flat budget across types so paragraph-length project memories
+    # aren't starved by an even slice. Types are still walked in order
+    # (identity/preference first when they're the input), so earlier
+    # types still get first pick when the budget is tight.
+    for mtype in memory_types:
         for mem in get_memories(
             memory_type=mtype,
             project=project,
@@ -381,11 +384,10 @@ def get_memories_for_context(
         ):
             entry = f"[{mem.memory_type}] {mem.content}"
             entry_len = len(entry) + 1
-            if entry_len > type_budget - type_used:
+            if entry_len > available - used:
                 continue
             selected_entries.append(entry)
-            type_used += entry_len
-        available -= type_used
+            used += entry_len
 
     if not selected_entries:
         return "", 0