Same verdict. Honest arithmetic.

Every gate in this report reads one universe — the broad NSE cash segment, n=2,967 — because the iList reads roughly seven points richer and a regime that flips on which source happened to load is not a regime. On that single gauge the picture is unambiguous: no cohort is above the 50% waterline, the healthiest is the 50-DMA at 43.7%, and the swing cohort at 31.4% is 13.6 points below where a tradeable tape begins. Against the index, Nifty 500 is up just 0.5% over three months — the first time this report has carried a benchmark at all, and it reframes every "strong" name below: outperformance here is a low bar cleared, not a bull market. Sector structure is narrow but real, with retail trade and health services the only cohorts printing meaningfully positive weeks.

Bias: Bull · 64.7% Trend: Uptrend · 62.7% Swing: Upswing · 50.5%  · 1 session(s) of history · India-calibrated thresholds applied verbatim — monitoring labels, not gates.

593 iList names cleared at least one of the eleven signals, each rescored 0–5 against the five published criteria in config/rules.yaml — the rubric is now written down rather than re-derived each week. 35 names score 5/5 and 230 clear the new relative-strength gate at the 70th percentile versus Nifty 500. The HMS filter eliminated 151 names outright at 0/3 (below EMA20, SMA50 and SMA200 simultaneously) and passed 171 at full alignment. Signals are computed live from the TradingView scanner this week, not from the external scan pack: the newest export on disk is 5-June vintage, and a seven-week-old pattern layer is now excluded from scoring by the freshness rule rather than quietly used.

Every name in this module is watch-only — the market gate above it is shut. The methodology changed materially this issue and the change is the point. Stops are no longer placed to hit a risk number; they are placed where volatility and structure require, at max(1.5 × ADR, support − 0.5%), and the resulting stop distance is an output. Risk is then capped by position size: every idea below is sized so the rupee loss at its stop equals 0.99% of capital, whether that stop sits 7.0% away or 9.9% away. The old rule — a 3% cap on the stop distance itself — was unsatisfiable against a 1.5×ADR stop on an ADR ≥ 3% name, and that is not a theoretical objection: the first run of this pipeline under the old rule passed 65 names through the liquidity stage and zero through risk. Circuit bands are now verified against live Dhan circuit limits rather than assumed, which removed 17 names sitting in 5% or 2% bands from consideration — precisely the trap the MOBAVENUE rule was written for.

All five picks carried from the 18-July issue were explicitly no-capital watch items, so this is a scoreboard of the framework's eye rather than the account. The basket is flat — an average of +0.1% — through a week in which swing breadth made another low, and flat is outperformance when the alternative was being long. Two did what leaders do in a correction and went up: HSCL +6.5% (now 3.7% off its 52-week high) and CONFIPET +7.1% (2.8% off). Two drifted. The instructive one is ACUTAAS at −9.94%: last issue's highest-conviction structure, and it now sits six hundredths of a percent above the −10% invalidation line. Read that honestly — the rule will call it "developing" for one more week on a rounding margin, which is exactly the kind of threshold that should never be the only thing standing between a thesis and a decision. Had Tranche-1 capital been in it at the listed support, the closing-basis stop would have cut it near −3% rather than −10%. The gate is what made the difference, not the chart.