Sydnee nightly — EXECUTION_QUALITY audit — 2026-08-14 P0 findings: 0 P1 findings: 10 Risks: 2 - Area: Execution Quality (Friday theme) - Branch: `dev` (3c9ec34 — docs(peak-monitor): hourly log 2026-08-13 13:15 PT) - Files scanned: `bot.py` (24,173 lines), `core/risk.py` (150 lines), `docs/audit_2026-08-07_EXECUTION_QUALITY.md`, `git log --oneline -30` - **Bugs found (P0 / P1): 0 / 10 (all carried from 2026-08-07 — 0 new)** - **Risks noted: 2 (both carried from 2026-08-07)** **Cross-check:** `git log --oneline -30` shows zero commits to `bot.py`, `core/`, or any strategy/core file since 2026-07-17. All 10 P1 bugs and 2 risks from the 2026-08-07 EXECUTION QUALITY audit re-verified at current line numbers — all present and unchanged. `_kcr_entry_ms` P1 (filed 2026-06-26) confirmed RESOLVED via the `datetime.fromisoformat` fallback at `bot.py:20742–20752` (comment references "Audit 2026-06-30 P1") — not carried, added to OK. --- Full report (dev branch): https://github.com/kanex1/sydnee.signals/blob/dev/docs/audit_2026-08-14_EXECUTION_QUALITY.md Reply FROM [email protected] to [email protected] to request fixes, e.g.: "code_task on sydnee-signals-dev: apply fix for the P0 about RVOL threshold in bot.py" Sydnee Agent will propose + you APPROVE (or plain 'approve') + auto-push to dev. --- Full audit below (first 12 KB) --- # Nightly Audit 2026-08-14 — EXECUTION QUALITY ## Summary - Area: Execution Quality (Friday theme) - Branch: `dev` (3c9ec34 — docs(peak-monitor): hourly log 2026-08-13 13:15 PT) - Files scanned: `bot.py` (24,173 lines), `core/risk.py` (150 lines), `docs/audit_2026-08-07_EXECUTION_QUALITY.md`, `git log --oneline -30` - **Bugs found (P0 / P1): 0 / 10 (all carried from 2026-08-07 — 0 new)** - **Risks noted: 2 (both carried from 2026-08-07)** **Cross-check:** `git log --oneline -30` shows zero commits to `bot.py`, `core/`, or any strategy/core file since 2026-07-17. All 10 P1 bugs and 2 risks from the 2026-08-07 EXECUTION QUALITY audit re-verified at current line numbers — all present and unchanged. `_kcr_entry_ms` P1 (filed 2026-06-26) confirmed RESOLVED via the `datetime.fromisoformat` fallback at `bot.py:20742–20752` (comment references "Audit 2026-06-30 P1") — not carried, added to OK. --- ## Findings ### BUG [P1] (carried from 2026-08-07): `Cancelled/Inactive` retry in `_execution_qa` missing `start_time` reset — retry order killed after ≤6 seconds **File:** `bot.py:7984` **Evidence:** ```python # bot.py:7975-7985 — Cancelled/Inactive retry path if not retried: retried = True ... exec_trade = self._place_order_with_overnight_fallback(_ct2, order2, ss.contract) oid = exec_trade.order.orderId self._active_order_ids.add(oid) continue # ← no start_time = time.time() before this # 30s fallback (bot.py:8037-8050) and PreSubmitted block (bot.py:8030-8034) # both correctly reset start_time with identical comments. ``` `_place_order_with_overnight_fallback` can consume 2–3s. Without `start_time` reset, the retry order inherits the near-exhausted original budget. Worst case: cancel at t=113s, new order placed at t≈116s, while-loop exits at t=119s, retry order killed after only ~3s of fill monitoring. An IOC entry that fills during those 3s is undetected, creating an orphaned IBKR position. **Impact:** `Cancelled`/`Inactive` retry orders get ≤6s of fill monitoring instead of a full 120s window. On active market hours with IOC on thin-book names, retry is guaranteed to be killed. Entry silently dropped with a misleading "EXEC QA: TIMEOUT 120s" log that obscures the budget-exhaustion root cause. **Fix:** Add `start_time = time.time()` at `bot.py:7985`, after `self._active_order_ids.add(oid)` and before `continue`. --- ### BUG [P1] (carried from 2026-07-10 / 2026-08-01 / 2026-08-07): `OBV LOSS EXIT` fires on profitable trades below cushion threshold — winners force-exited at market **File:** `bot.py:5114` **Evidence:** ```python # bot.py:5113-5128 _min_cushion = max(float(t.entry_price) * _cushion_pct, atr_val * 0.25) if _profit_guard and profit_per_share < _min_cushion: # _min_cushion always > 0 ... if _loss_exit: hit = True reason = ... "at loss after ..." # ← misleading: could be small winner self._log_activity(sym, "OBV LOSS EXIT: ... at loss ...", "trade") ``` `< _min_cushion` is always a positive threshold (e.g. $0.75/sh on a $100 stock with $3 ATR). Profitable trades with small gains below the cushion satisfy the condition and get force-exited at market with a false "at loss" log entry. The `_loss_exit` inner guard is necessary but not sufficient — `profit_per_share < _min_cushion` is the wrong predicate. **Impact:** On dev (`OBV_LOSS_EXIT=true`): winners held past 5 bars with a small favourable move are ejected before profit target. Prod safe (flag default off). **Fix:** ```python if _loss_exit and profit_per_share < 0: hit = True ... ``` --- ### BUG [P1] (carried from 2026-07-04 / 2026-08-07): `risk.state.trade_count` not restored from DB on restart — `max_daily_trades` global cap bypassed after every restart **File:** `core/risk.py:118`, `bot.py:23457–23460` **Evidence:** ```python # core/risk.py:30-34 — DailyState always starts zeroed: @dataclass class DailyState: realized_pnl: float = 0.0 trade_count: int = 0 # bot.py:23457-23460 — startup only calls set_open_positions(): open_count = len(self.open_trades()) if open_count > 0: self.risk.set_open_positions(open_count) # No restore of trade_count or realized_pnl # core/risk.py:118 — cap that relies on trade_count: if self.state.trade_count >= self.cfg.max_daily_trades: return False, "Max daily trades reached" ``` `RiskManager` has no `restore_daily_state()` method. Every mid-session restart resets `trade_count` to 0 and `realized_pnl` to 0.0. A bot at the max daily trade count that restarts places new trades freely; a bot past the daily loss limit bypasses the circuit breaker for the rest of the session. **Impact:** Real-money risk if `max_daily_trades` or `daily_loss_limit` are actively managing exposure. Both caps are silently nullified after any restart. **Fix:** At startup, query `SELECT COUNT(*), SUM(pnl) FROM trades WHERE trade_date = today AND exit_time IS NOT NULL` and restore via `self.risk.record_exit(pnl)` per closed trade, or add `restore_daily_state(trade_count, realized_pnl)` to `RiskManager`. --- ### BUG [P1] (carried from 2026-07-03 / 2026-08-07): `_execution_qa` `IndexError` when `"Filled"` arrives before fill detail — orphaned IBKR position **File:** `bot.py:7951` **Evidence:** ```python # bot.py:7946-7951 if status == "Filled" or (exec_trade.fills and filled > 0): fp = float(exec_trade.orderStatus.avgFillPrice or 0) or \ float(exec_trade.fills[0].execution.price) # ^^^ IndexError if fills == [] and avgFillPrice == 0 ``` When IBKR fires `"Filled"` status before `execDetails` populates the `fills` list, `fills[0]` raises `IndexError`. No `Trade` record is created → orphaned IBKR position untracked in DB. **Impact:** Race is rare but reproducible on fast MKT fills in liquid markets. Results in IBKR holding shares with no bot-side trade record; reconcile re-imports as an untracked position, distorting risk state. **Fix:** ```python fp = float(exec_trade.orderStatus.avgFillPrice or 0) if not fp: if exec_trade.fills: fp = float(exec_trade.fills[0].execution.price) else: continue # fills not yet populated — wait one more poll ``` --- ### BUG [P1] (carried from 2026-07-03 / 2026-08-07): `broker_executions.timestamp_sent` always equals `timestamp_filled` — order-to-fill latency unmeasurable **File:** `bot.py:7910` **Evidence:** ```python # bot.py:7899-7911 ts_filled = _dt.datetime.now(_dt.timezone.utc) ... VALUES (%s,%s,...,%s,%s) (... ts_filled, ts_filled)) # ← both timestamp_sent and timestamp_filled = now() ``` `_persist_broker_execution` does not accept `start_time` as a parameter. `start_time` is available in `_execution_qa` (defined at `bot.py:7934`) but never threaded in. `timestamp_filled − timestamp_sent` is always 0 ms for every row. **Impact:** `broker_executions` latency column is permanently useless. Execution degradation (slow fills during high-VIX events, thin pre-market) is invisible. **Fix:** Add `order_start_time` param to `_persist_broker_execution`; compute `ts_sent = datetime.fromtimestamp(order_start_time, tz=UTC)` at the top of the function. Pass `start_time` at the call site (bot.py:7960). --- ### BUG [P1] (carried from 2026-07-03 / 2026-08-07): `signals.tempo` uses P&L/share as execution-quality proxy — actual slippage never queried **File:** `bot.py:11147` **Evidence:** ```python # bot.py:11145-11148 # Per-share P&L as proxy for execution quality per_share = t.pnl / t.shares if t.shares > 0 else 0 slippages.append(per_share) # ← variable name misleading; not slippage hourly_slippage.setdefault(entry_hour, []).append(per_share) ``` `broker_executions.slippage` is never read by `_tempo_daily_analysis`. Hourly breakdown reflects trade profitability, not execution degradation. Bad fills (retried orders, aggressive-limit slippage) are invisible in the tempo output. **Impact:** Execution quality signal is blind to the very metric it claims to track. Slow-fill hours and venue degradation events go undetected in the tempo report. **Fix:** In `_tempo_daily_analysis`, join `broker_executions` on `symbol + timestamp_filled` grouped by hour; report `avg(slippage)` separately from P&L. --- ### BUG [P1] (carried since 2026-06-12 / 2026-08-07): `config_syms` NameError in `_exec_qa_patrol` — position mismatch check silently never runs **File:** `bot.py:8333` **Evidence:** ```python # bot.py:8272: watchlist_syms = set(self.symbols) ← correct variable defined here # bot.py:8333: if ib_qty != bot_qty and sym in config_syms: # ← NameError: config_syms undefined ``` `config_syms` is undefined in `_exec_qa_patrol` and at class scope. The surrounding `except Exception: pass` (bot.py:8336) swallows the `NameError` on every patrol call. IBKR/bot position discrepancies are never detected. **Impact:** Check 5 of `_exec_qa_patrol` has never executed since this code shipped. Orphaned IBKR positions (from outside-bot trades, stale reconcile imports, or the IndexError P1 above) go undetected intraday. **Fix:** `bot.py:8333` — replace `config_syms` with `watchlist_syms`. --- ### BUG [P1] (carried since 2026-06-12 / 2026-08-07): `_execution_qa` 30s MKT fallback fires outside RTH — cancels valid resting LMT, places rejected MKT, silently drops entry **File:** `bot.py:8037` **Evidence:** ```python # bot.py:8037 if elapsed > RETRY_AT and not retried and not is_manual: # no RTH guard retried = True self.ib.cancelOrder(exec_trade.order) # ← cancels valid resting off-hours LMT ... order2 = MarketOrder(action, shares) # ← IBKR rejects MKT outside RTH ... start_time = time.time() # budget reset present, but RTH guard missing ``` No `and in_rth` check. Off-hours resting LMT cancelled after 30s and replaced with a GTC-wrapped MKT that IBKR rejects. `retried=True` → next rejection → handler returns `(None, None)`. Entry is silently abandoned. **Impact:** `rsi_extreme` pre/post-market entries that don't fill on the first liquidity touch are abandoned after ~33s rather than resting in the off-hours book. The 24×5 strategy edge is clipped every time a pre/post-market LMT takes more than 30s to fill. **Fix:** Add RTH guard at `bot.py:8037`: ```python if elapsed > RETRY_AT and not retried and not is_manual and in_rth: ``` Outside RTH: refresh the aggressive limit price via `_get_aggressive_limit` and re-place a new LMT instead. --- ### BUG [P1] (carried since 2026-06-12 / 2026-08-07): `RSI_TICK_FIRE` window ends at 780 (13:00 ET) — three peak RTH afternoon hours cut **File:** `bot.py:22189` **Evidence:** ```python # bot.py:22187-22189 if not (570 <= t_min < 780) or now_et.weekday() >= 5: # Not RTH — reset state and skip ``` `780 = 13:00 ET`. RTH closes at `960 = 16:00 ET`. `rsi_extreme` entries and the tick-fire state machine are blocked from 13:00–16:00 ET when `RSI_TICK_FIRE=true`. The 13:00 ET bucket yielded +0.534% avg / 67.2% WR in the 2026-05-19 backtest — the single highest-edge slot. **Impact:** Three RTH hours of `rsi_extreme` signal completely suppressed. `RSI_TICK_FIRE=true` is the default on dev; same default applies if promoted to production. **Fix:** `bot.py:22189` — change `780` → `960`: ```python if not (570 <= t_min < 960) or now_et.weekday() >= 5: ``` --- ### BUG [P1] (carried since 2026-06-19 / 2026-08-07): `_tape_last_loss_ms` in-memory only — TAPE loss cooldown resets on every bot restart **File:** `bot.py:21312–21314` (write), `bot.py:21707–21709` (read) **Evidence:** ```python # bot.py:21312-21314 — write on -1% stop hit if not hasattr(self, "_tape_last_loss_ms"): self._tape_last_loss_ms = {} self._tape_last_loss_ms[sym] = now_ms # ← dict in memory, no Redis write # bot.py:21707-21709 — gate at entry if not hasattr(self, "_tape_last_loss_ms"): self._tape_last_loss_ms = {} _last_loss = self._tape_last_loss_ms.get(sym, 0) # ← always 0 after restart ``` No Redis write, no DB write, no restore path. `TAPE_LOSS_COOLDOWN_MIN` gate effe