feat(routing): wire AI target_asset end-to-end — bot trades the right perp

Closes the loop on the asset-routing prompt change. Previously the v5
prompt emitted target_asset (e.g. SOL, TRUMP) but bot_engine still
read price_impact_asset and only ever traded BTC/ETH. Now the trade
actually fires on whatever perp the AI picked.

Schema (alembic 006):
  posts.target_asset       (str)   — HL perp ticker, any of the universe
  posts.category           (str)   — 6-class enum (direct_named, etc.)
  posts.expected_move_pct  (float) — AI's 1h move estimate

Wiring:
  truth_social.py persists the three fields when creating Post rows.
  bot_engine.py routing:  asset = target_asset || price_impact_asset || BTC
  Old rows (target_asset=NULL) fall back to legacy BTC/ETH path — no
  retroactive scoring needed; new rows route correctly from now on.

Hyperliquid trader doesn't need changes — `coin` is already a parameter,
and analysis.py validated against HL_PERPS before storing target_asset
so by the time bot_engine reads the field, it's guaranteed tradeable.

Deployment:
  alembic upgrade head    # adds the 3 columns
  Restart api container

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

alembic/versions/006_add_target_asset.py

@@ -0,0 +1,34 @@
+"""Add target_asset / category / expected_move_pct to posts (v5 routing)
+
+Revision ID: 006
+Revises: 005
+Create Date: 2026-05-08 00:00:00.000000
+
+The v5 AI prompt outputs not just signal direction but also which specific
+perp to trade (could be BTC/ETH/SOL/TRUMP/anything on Hyperliquid). The
+bot reads `target_asset` to route the trade. `price_impact_asset` stays
+bound to BTC/ETH for the existing price_impact_monitor.
+"""
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+
+revision: str = "006"
+down_revision: Union[str, None] = "005"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    with op.batch_alter_table("posts") as batch_op:
+        batch_op.add_column(sa.Column("target_asset", sa.String(16), nullable=True))
+        batch_op.add_column(sa.Column("category", sa.String(24), nullable=True))
+        batch_op.add_column(sa.Column("expected_move_pct", sa.Float(), nullable=True))
+
+
+def downgrade() -> None:
+    with op.batch_alter_table("posts") as batch_op:
+        batch_op.drop_column("expected_move_pct")
+        batch_op.drop_column("category")
+        batch_op.drop_column("target_asset")

app/models.py

@@ -53,6 +53,13 @@ class Post(Base):
     ai_reasoning: Mapped[Optional[str]] = mapped_column(Text, nullable=True)
     prefilter_reason: Mapped[Optional[str]] = mapped_column(String(32), nullable=True)
     analysis_version: Mapped[Optional[str]] = mapped_column(String(16), nullable=True)
+    # v5: AI-decided trade target. Can be any HL perp (BTC/ETH/SOL/TRUMP/...),
+    # not just BTC/ETH. The bot routes the actual trade here.
+    # `price_impact_asset` stays bound to BTC/ETH for the existing
+    # price_impact_monitor; `target_asset` is what we actually trade.
+    target_asset: Mapped[Optional[str]] = mapped_column(String(16), nullable=True)
+    category: Mapped[Optional[str]] = mapped_column(String(24), nullable=True)
+    expected_move_pct: Mapped[Optional[float]] = mapped_column(Float, nullable=True)
     created_at: Mapped[datetime] = mapped_column(DateTime, nullable=False, default=utcnow)
 
     trades: Mapped[List["BotTrade"]] = relationship("BotTrade", back_populates="trigger_post")

app/scrapers/truth_social.py

@@ -102,11 +102,16 @@ async def _process_entry(entry: dict, db: AsyncSession) -> Optional[Post]:
         prefilter_reason=analysis.get("prefilter_reason"),
         analysis_version=analysis.get("analysis_version"),
         relevant=analysis["relevant"],
+        # `asset` (BTC/ETH only) feeds the existing price_impact tracker.
         price_impact_asset=asset if analysis["relevant"] else None,
         price_impact_m5=None,    # filled by price_impact_monitor after 5 m
         price_impact_m15=None,   # filled by price_impact_monitor after 15 m
         price_impact_m1h=None,   # filled by price_impact_monitor after 1 h
         price_at_post=price_at_post,
+        # v5 routing: AI decides the actual perp to trade. May be SOL/TRUMP/etc.
+        target_asset=analysis.get("target_asset"),
+        category=analysis.get("category"),
+        expected_move_pct=analysis.get("expected_move_pct"),
     )
     db.add(post)
     await db.flush()

app/services/bot_engine.py

@@ -77,8 +77,17 @@ async def process_post(post: Post, db: AsyncSession) -> None:
         logger.info("Post %d skipped: signal=%s is not actionable", post.id, post.signal)
         return
 
-    asset = post.price_impact_asset or 'BTC'
+    # v5: route the trade to the AI-decided perp. target_asset can be any
+    # Hyperliquid perp ticker (BTC/ETH/SOL/TRUMP/...) — analysis.py already
+    # validated against HL_PERPS and resolved chain fallbacks before this
+    # field was set, so by the time we get here it's safe to trade directly.
+    # Fallbacks for older / null rows: legacy price_impact_asset, then BTC.
+    asset = post.target_asset or post.price_impact_asset or 'BTC'
     side = 'long' if post.signal == 'buy' else 'short'
+    logger.info(
+        "Routing post %d → trade %s/%s (category=%s, expected_move=%.2f%%)",
+        post.id, asset, side, post.category, post.expected_move_pct or 0,
+    )
 
     logger.info("Signal: post=%d asset=%s side=%s confidence=%d", post.id, asset, side, post.ai_confidence)