Cut database egress 189x by deferring the prompt snapshot

The free-tier 5 GB/month network transfer allowance ran out, which blocks
connections outright. The database is only ~55 MB, so 5 GB meant the whole
thing was being pulled roughly 90 times over.

Cause: actions is 39 MB of that 55 MB -- 541 rows at ~74 KB each, almost
entirely context_snapshot, which stores the whole assembled prompt for a
turn. Every adventure load and every turn fetched all of it in order to
read two small things out of it: the world-change chips under an AI
message (Action.world_changes) and the emit block re-attached when
replaying history to the model (_history_text). The Insights viewer is
the only consumer that wants the whole snapshot, and it asks for one
action at a time.

Lifts that slice into its own small actions.world_delta column
(migration 36) and marks context_snapshot, state_before and
world_state_before deferred, so they load only when something touches
the attribute -- Insights, undo and retry, all single-action paths.
The backfill runs server-side, dialect-specific (json_extract on SQLite,
#> on Postgres), because pulling 39 MB of snapshots into Python to
rewrite a slice of each would defeat the purpose.

Measured at production shape (541 actions, 72 KB snapshots), one
adventure load goes from 38.46 MB to 0.20 MB. The traffic that consumed
5 GB would now be about 27 MB.

Deliberately not included: limiting the history query to recent actions,
and removing the redundant db.refresh(adventure) calls. Both were sized
against the old numbers; against a 0.20 MB load they would take ~27 MB a
month down to ~10 MB, which is not worth the complexity.

tests/test_egress.py hooks before_cursor_execute and asserts the emitted
SQL never names the deferred columns during a bulk load, so this cannot
regress silently. 123 tests pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UeQVy5bEjLhfgWNc27Efet

backend/app/context/builder.py

@@ -62,15 +62,17 @@ def _history_text(action: models.Action) -> str:
     delta). The block is stripped before storage/UI, so without this every past
     AI turn would look like one that emitted nothing — biasing the model, by
     imitation, to stop emitting too. Player turns and blockless turns are
-    returned unchanged."""
+    returned unchanged.
+
+    Reads `world_delta`, not `context_snapshot`: this runs for every action in
+    the replayed history, and the snapshot is deferred precisely so a turn
+    never drags the prompt archive out of the database."""
     text = action.text
-    snap = action.context_snapshot if isinstance(action.context_snapshot, dict) else None
-    if snap:
-        ws = snap.get("world_state")
-        if isinstance(ws, dict):
-            block = worldstate.render_delta_block(ws.get("delta") or {})
-            if block:
-                text = f"{text}\n{block}"
+    wd = action.world_delta if isinstance(action.world_delta, dict) else None
+    if wd:
+        block = worldstate.render_delta_block(wd.get("delta") or {})
+        if block:
+            text = f"{text}\n{block}"
     return text