feat(meet-kevin): prompt v2 — forward-looking action + expected_move field

User reported that the old prompt could emit 'sell' on backward-looking
capitulation ('Kevin sold after a 20% drop') — exactly the false signal
to avoid. v2 reframes every per-ticker field as forward-looking and
adds an explicit expected_move enum for the trading bot to weight.

Changes:
- New ExpectedMove enum (up_strong/up_mild/sideways/down_mild/
  down_strong/unknown) in shared/schemas + shared/models, with
  matching kevin_expected_move Postgres enum + column on
  kevin_stock_mentions (migration e5f6a7b8c9d0). NOT NULL with
  server_default 'unknown' so existing rows backfill cleanly.
- SYSTEM_PROMPT rewritten: action semantics now require a FORWARD
  view; reactive sells get downgraded to 'watch' or skipped; the
  rationale_quote must contain forward reasoning. Quality
  checklist updated.
- _ANALYSIS_TOOL JSON schema gains expected_move (required).
- prompt_version v1 → v2 in config + infra + ad-hoc CLI default.
- pipeline.py persists ticker.expected_move into the new column.

Migration safety: the column is NOT NULL DEFAULT 'unknown' so 96
existing mentions auto-fill with 'unknown' (no forward call known
for backward analyses) without breaking any reads.

Cost to backfill the 27 existing analyses with v2 prompt: ~$3 LLM
spend. A follow-up reanalyze script will replay them after this
ships.

scripts/analyze_kevin_video.py

@@ -10,7 +10,7 @@ no Redis publish — strictly observational.
 Env vars (or pass via flags):
     TRADING_ANTHROPIC_OAUTH_TOKEN   — required (matches what the pod uses)
     TRADING_MEET_KEVIN_LLM_MODEL    — default: claude-haiku-4-5-20251001
-    TRADING_MEET_KEVIN_PROMPT_VERSION — default: v1
+    TRADING_MEET_KEVIN_PROMPT_VERSION — default: v2
 """
 
 from __future__ import annotations
@@ -94,8 +94,8 @@ def print_analysis(video_id: str, captions, result) -> None:
         return
 
     print(f"\n  TICKERS ({len(a.tickers)}):")
-    print(f"    {'SYMBOL':<8} {'ACTION':<7} {'CONV':>6} {'HORIZON':<11} {'RATIONALE'}")
-    print(f"    {'-'*8} {'-'*7} {'-'*6} {'-'*11} {'-'*60}")
+    print(f"    {'SYMBOL':<8} {'ACTION':<7} {'EXPECTED':<12} {'CONV':>6} {'HORIZON':<11} {'RATIONALE'}")
+    print(f"    {'-'*8} {'-'*7} {'-'*12} {'-'*6} {'-'*11} {'-'*60}")
     # sort by action priority (buy/sell first), then conviction desc
     action_rank = {"buy": 0, "sell": 1, "hold": 2, "watch": 3, "avoid": 4}
     sorted_t = sorted(
@@ -107,9 +107,12 @@ def print_analysis(video_id: str, captions, result) -> None:
     )
     for t in sorted_t:
         rationale = (t.rationale_quote or "").replace("\n", " ")[:60]
+        expected = getattr(t, "expected_move", None)
+        expected_str = expected.value if expected is not None else "—"
         print(
             f"    {t.symbol:<8} "
             f"{t.action.value:<7} "
+            f"{expected_str:<12} "
             f"{fmt_pct(t.conviction):>6} "
             f"{t.time_horizon.value:<11} "
             f"{rationale}"
@@ -127,7 +130,7 @@ async def main() -> None:
     )
     parser.add_argument(
         "--prompt-version",
-        default=os.environ.get("TRADING_MEET_KEVIN_PROMPT_VERSION", "v1"),
+        default=os.environ.get("TRADING_MEET_KEVIN_PROMPT_VERSION", "v2"),
     )
     args = parser.parse_args()