feat(sl-hunting): v5f, SLH-017, SLH-018 and the 17 SEP pre-open note

[DoRmAmMu1997]

Four commits: the knowledge rule for this session, the two code gates for the arms of it that had to stop being prose, and tomorrow's pre-open note.

Same note, same morning, opposite branch

Source: TXiU5Odet04, IH's live session, uploaded 2026-09-16 10:09 IST.

The pre-open note shipped for today (PR #171) was conditional: gap up → BUY the trapped sellers, flat or gap down → SELL with the market. NIFTY opened +82 points (+0.36%), BankNIFTY +0.38%.

	Read of the open	Side	Outcome
IH	FLAT	bought puts, all three indices	booked ~his 3-lakh target
Agent	gap up	4 entries, all LONG	−₹863.75

IH: "today we saw a FLAT opening, and with this kind of opening we are going to follow the CONTINUED SELLING… if a VERY BIG gap up had opened we could not have worked like this."

Agent: "NIFTY gapped +82pts (0.36%) and BankNIFTY +213pts (0.38%) above prior close after a 457pt breakdown day that seated positional sellers, satisfying the pre-open note's GAP-UP→BUY branch."

The proximate cause is a v4t violation, not a knowledge gap

v4t already says the classification is a judgement you make and state, already calibrates it — "a quarter of a percent is not a gap… something in the region of half a percent is where the gap reading starts to earn itself" — and already says hesitation resolves to flat. 0.36% is below that line. The number was computed and written into all four entry rationales and never once tested against the rule.

This is the third session decided by that one word:

	What happened	Cost
31 Aug	read "gap-down" where IH read flat	produced v4t
01 Sep	measured against the official close, not the 3:15 level	−₹1,488.25 → produced v4u
16 Sep	read 0.36% as a gap-up	−₹863.75

Nothing downstream was wrong on any of the three.

What v5f actually encodes

Not the branch — the premise underneath it, which is why the agent wanted the gap-up branch at all. All four rationales asserted the 457-point breakdown had "seated positional sellers". IH argues the opposite, with a mechanism:

"when do sellers REMAIN seated? When a retracement happens. When there is no retracement, every trader keeps fearing that a retracement might come, that his trade might go wrong"

— so on continuous momentum "traders do come, but they also book their target and leave", and "the chances of HOLDING are low here". A violent one-way session ends with almost nobody carrying inventory, however large its range.

Then the falsification test, which is the part that runs before the trade:

"but if sellers had been seated in good quantity, what would the market have opened as? We would have seen a straight gap up."

The open is the receipt for the premise. A small gap is not a weak version of that open — it is the evidence against it.

A ONE-WAY DAY LEAVES NOBODY SEATED, AND THE NEXT OPEN IS THE RECEIPT (v5f) sits directly after the whole v4t block — including its v4u sub-bullet, which an earlier attempt split and orphaned before I caught it — so the model reads "decide which word describes the open", then "here is what the open's size is measuring". It is v5c one scale up: there a retracement recruits a crowd inside the move, here a retracement is what lets a crowd survive into the next session.

Verification

Negative-tested 14 ways, all 14 caught, plus a control mutation of unasserted prose that correctly did not trip the test. One anchor initially matched zero times because it began mid-line rather than at a line start; re-run line-contiguously, it caught.

Gates: 606 master · 28 market-data-health · 1508 pytest (+12) · ruff 0.16.7 · mypy (80 files + master) · compileall · bandit · coverage 73.8%, policy pass.

Considered and not encoded

• "A retracement is what seats a crowd" — partly v5c already; only the cross-session half is new, and that is what v5f carries.
• His loss-discipline material — repeats v3y.
• His round-number caution — already covered in BNF_SPECIFIC.

Commit 2 — SLH-017, the open is classified in code

The arm of this session that had to stop being prose. pivot_and_levels now returns an open_classification block:

previous_close / today_open / gap_points / gap_pct
threshold_pct  (GAP_CLASSIFICATION_THRESHOLD_PCT = 0.5)
verdict        FLAT | GAP_UP | GAP_DOWN
reference      "previous session's last candle close, not the official 15:30 close"

Neither number is invented. The 0.5% threshold is v4t's own calibration; sub-threshold opens read FLAT in both directions because v4t says hesitation resolves to flat. The reference is v4u's, for v4u's stated reason — the closing auction carries settlement prints no crowd traded around.

It reaches the model through levels and, for BankNIFTY, through bank_nifty. Run against 16 Sep's real numbers, both come back FLAT (NIFTY 0.355%, BankNIFTY 0.381%) — IH's reading of that morning, and the opposite of the one the agent acted on.

Wired, not merely added. v4t now carries an SLH-017 paragraph telling the model to read the verdict and state it, and not to recompute or argue past it — an unwired fact is one the model re-derives its own way. The existing v4t marker test was extended, not rewritten, so its original assertions still pin the original prose.

Scope kept narrow: this supplies the arithmetic and the threshold. It rejects no orders and does not judge what the classification means — that stays with the model.

TDD: six tests RED first, then the implementation. Negative-tested 9 ways, all 9 caught, plus a passing control. One mutation — measuring from today's LAST price instead of its OPEN — initially passed, because every fixture opened and closed at the same price. That was a real defect the tests could not see: a flat open would silently become a "gap up" by mid-morning and re-select the branch hours after the fact. A dedicated drift test was added; the mutation is now caught.

Commit 3 — SLH-018, one round of friction on a reasonless EXIT

The session's other item: an EXIT with no reason closed the day's only winner (+₹1,182), about 100 seconds before that trade's 23276 target printed.

I raised this as a gate and then declined to build it, because refusing such an exit reverses a standing decision:

• SLH-007: an EXIT is never refused over its wording; bouncing it would strand an open position.
• SLH-010 explicitly declined to reverse SLH-007 and added the warning instead.
• test_exit_with_no_reason_still_executes_but_warns_loudly existed so "a later tightening cannot quietly turn the warning into a refusal".

The operator then took the decision explicitly, choosing the one-round form — it "would filter out the anomalies and let SLH-007 work on genuine wordless exit cases". SLH-007 is narrowed, not reversed, and the tightening is neither quiet nor unilateral.

The contract. The first EXIT resolving to the no-reason sentinel is sent back once, with a rejection naming both ways forward — state the reason, or send the EXIT again. Any repeat is honoured unconditionally whatever it carries. Worst case is one decision cycle, never a stranded position; stop/target/max-loss/square-off are host-owned throughout and never gated.

Where the state lives matters. The armed flag sits on the executor, not the tool context. A context is rebuilt every bar, so context-local state would bounce the first attempt of every bar forever and could genuinely strand a position — the exact outcome SLH-007 forbids. Set and cleared duck-typed under contextlib.suppress, since TradeExecutor is a Protocol.

Scoped per episode: any accepted order clears the arm, so a bounce spent on this morning's slip cannot buy a later position a free pass.

Placeholder reasons bounce too — _safe_exit_reason maps "placeholder"/"n/a" to the same sentinel, so they are the same class of anomaly. That guard test was rewritten to the two-step contract, keeping its real assertions.

The tool description had to change, or the model keeps a false contract. It said an EXIT "is NEVER rejected". It now states the single exception and the way out, and the test asserts the stale absolute is gone rather than merely that the new wording is present — otherwise the model holds a contract its tool does not keep, which is the SLH-017 unwired-fact failure arriving from the other side.

TDD: five tests RED first. Negative-tested 7 ways, all 7 caught, plus a passing control. The two that matter most are both caught: removing the bounce, and never disarming (which would strand a position by refusing the second attempt too).

Commit 4 — the pre-open note for 17 Sep

From SQIo4PisLw0, uploaded 2026-09-16 21:25 IST.

The branch shape repeats 16 Sep's note — gap up → BUY, flat or gap down → SELL — so what the test pins is the different reasoning underneath. Today's selling broke down and then recovered without crossing the higher point or the round number, which leaves one question open: which side did the session trap?

Open	Side	Why
Flat / gap down	SELL	the sellers were only trapped, never seated — "seeing the retracement, who comes here? SELLERS come. But because the market stays SIDEWAYS they leave again"
Gap up	BUY	a lone deliberate buyer may be — "it did not cross the round number, it did not cross its high point", so "maybe some good trader slowly bought here and other people could not buy"

Because "the higher point was not crossed, retail did not do much work", only a deliberate buyer may be seated, never a crowd — which is exactly why that branch needs the gap up to confirm one exists rather than assuming it. SENSEX expiry is flagged twice, as context for the triple-index read, not a premise.

NIFTY      R 23420 / 23280   S 23110 / 23000
BANKNIFTY  R 56300 / 56650   S 55810 / 55500
SENSEX     R 75000 / 74500   S 74000 / 73650

Three levels needed operator confirmation against the video. NIFTY's first support arrived as 2310 and is 23110 — a dropped digit, not a dropped trailing zero, so my 23100 reading was wrong. BankNIFTY's second support arrived as 5500 again and is 55500 again. SENSEX was the worst yet: both pairs came through as single runs, 7574500 and 7473650, resolving to round thousands — 75000/74500 and 74000/73650 — which neither reconstruction I offered had guessed. Recorded in the test comments so the next garble of this shape reads round numbers first.

Negative-tested 18 ways, all 18 caught, plus a passing control — including guards on all three corrected levels, so a silent revert to my wrong readings fails.

No doc section: notes haven't carried one (PR #168 precedent).

🤖 Generated with Claude Code