BTC

Analysis complete. Record persisted to the digests convention.

BTC/USDT · 1h · MOMENTUM BREAKOUT → WAIT (score 12/100)

Strategy selected via the real registry extension (loaded .pi/extensions/strategy-registry.ts and invoked its actual select-strategy handler — success:true, 7 strategies registered, weights confirmed 40/35/10/15). It isn't in my native tool set, so I drove it through the extension itself rather than faking the call.

All four entry conditions fail

Gate	Result
Score ≥ 70	✗ 12
RSI > 50 & rising	✗ 37.98, and last tick down (38.07→37.98); 10-bar slope −0.10
Volume > 20-avg	✗ 0.48× (1,144 vs 2,397)
HTF aligned	✗ 4h bearish — RSI 31.9, price < EMA20 < EMA50
Key level broken	✗ −2.21% below the 20-bar high; no open bullish FVG exists

Price sits inside the prior-day range (16.6%) and at just 9.1% of the prior-week range, below a bearish EMA20/50/200 stack, under descending swing highs (79,744 → 78,532 → 77,515). The forming bar pushing to ~77,100 is mitigating into overhead bearish FVGs from below — a retracement, not a breakout. Volume fails even pace-adjusted (0.92× projected).

Two findings worth your attention

1. The stored 4h/15m/5m volume data is corrupt. stored_4h_vol / Σ(1h vol) has a median ratio of 0.0117 (range 0.0000–10.31); 15m and 5m carry 330 and 980 zero-volume bars, with 15m median 0.93 vs 1h median 1,098.9. Any HTF or 15m volume reading from those files is meaningless — the tempting "15m vol/avg20 = 0.00" and "4h vol = 1.64×" are pure artifacts. I rebuilt 4h and 1d from the 1h series, which is clean (0 gaps, 0 non-positive volumes). This is worth fixing upstream, since it silently feeds any agent using those timeframes.

2. The strategy has no short-horizon edge on this instrument. Backtesting the rules as coded over 2,849 bars at 20 bps: long breakouts (n=106) return −0.125%/4h, 27.4% win, with negative medians out to 16h; short breakdowns (n=117) are worse. And the current-state analog (n=191) — low-volume bounce under a bearish stack — averaged −0.26%/4h to −0.52%/24h, 28.8% win; the last occurrence (Sep-10 00:00) resolved down into 76,394. So this isn't merely "setup not formed yet"; the tape I'd be buying has historically been a trap.

Honest caveats

Sentiment (0/10) and Risk (0/15) scored zero because those tools were unreachable here, not because they read badly — I did not infer them. Sensitivity check: crediting both unknowns at 50% still yields 24.5/100, far below the CONFIRM bar, so the WAIT verdict doesn't hinge on the gaps.

Independent of the score, ENTER was never available: no account equity/margin, risk-manager state unreachable, and execution-check/risk-gate/size-position/position_tracker all uncalled. Per the production hard stops, that's a mandatory WAIT with missing gates listed.

Watch plan (conditional, not an order): long only becomes scoreable on a closed 1h bar above 77,514.8 → 77,991 (trendline) → 78,503.5, with volume > 20-avg, RSI through 50, and a 4h close back over 78,070. Invalidation on a close below 76,394.1, which opens the 76,257.4 weekly low.