[eaitbrahim]

Closes #960
Closes #961

What changes

#961: one bracket verdict, read by three surfaces.

• status.bracket_verdict is now the only Paper positions render an unactionable 'NO bracket' warning — the state paper can never leave #641 judgement. These all read it:
  • keel status (_bracket_display);
  • the bracket cell on Pulse's roll-call and on Holdings (payload._bracket_field);
  • Pulse's missing-bracket count (pulse.build_pulse).
• A non-paper tranche is NEUTRAL "no stop by design" when both of these hold:
  • its rule kind declares that it never brackets. This is Rule.places_bracket = False, a class attribute declared once, on Dca;
  • it recorded no real stop.
• Every other non-paper tranche without a bracket stays WARN "NO bracket":
  • a DCA tranche that recorded a stop;
  • any non-DCA tranche, with or without a stop. The PAXG tranche (doctor: nothing warns when an open tranche has no live rule and no retry record -- PAXG sat unwatched for five cycles #811) is one of these;
  • a tranche whose kind keel does not know.
• Pulse's "brackets missing" count no longer counts DCA tranches.
• Paper is unchanged.
• Zero is not a stop. Every attributed live DCA tranche records initial_stop = 0, which is the Dca.detect stop=0 sentinel. The live ledger's tranches 4–6 show this. A check of initial_stop is None alone would have left all of them warning.
  • status.recorded_stop treats a stop ≤ 0 as no stop.
  • The Holdings stop column reads the same helper, so a "$0.00" stop no longer sits next to "no stop by design".

#960: Holdings draws one table per rule row.

• HoldingsReport.groups replaces rules. It is keyed on positions.rule_id, and each group's heading is kind · product · rule <id> (for example dca · BTC-USD · rule 6).
• Unattributed tranches come last, as one table per kind (dca · unattributed, turtle_breakout · unattributed).
• Group keys are built only from an integer or an ordinal (rule-6, unattributed-1). That makes the h-hold-<key> element ids unique and free of whitespace, which settles the earlier held H8.
• Each row carries group. The client filters on row.group === group.key and prints the service's heading.

Rulings

1. The heading uses the rule id, not cadence or amount.
   • Why: kind plus product does not tell same-product DCA rules apart (the paper ledger has rules 19 and 20, both BTC). Params can be edited and can coincide. The id is unique and is the handle keel rules takes.
   • Cost if wrong: the heading carries no budget or cadence; those are on Rules and Pulse.
2. Unattributed tranches are split by kind.
   • Why: one "unattributed" table would put pre-v21 DCA legs and the PAXG turtle tranche together with nothing on screen to show their kind. Grouping by name used to show it.
   • Cost if wrong: up to one extra table per kind.
3. A stop ≤ 0 counts as no stop, both for the verdict and for the Holdings stop column.
   • Why: it is the DCA sentinel, and a stop at zero protects nothing on a long.
   • Cost if wrong: a tranche whose recorded stop is really zero would read as having none. No real rule sizes against a zero stop.
4. places_bracket is a new class attribute; accumulates is not reused.
   • Why: the sleeve-sell kinds also accumulate, and they open no tranche at all.
   • Cost if wrong: a future stop-less entry kind has to declare it. test_dca_is_the_one_kind_that_never_brackets forces that to be a deliberate change.
5. Paper's bracket text is unchanged, including "paper stop 0" on paper DCA legs. The brief said paper behaviour stays as it is.

Verification

• Ran ruff check keel tests packages, ruff format --check . and bare mypy: all clean. node --check passes on every static JS file and on sw.js.
• pytest: 8286 passed, 3 skipped.
• Mutation testing: 15 of 15 mutants were killed.
  • The mutants covered: the verdict without the kind check, the zero sentinel counted as a stop, an unknown kind excused, the judgement keyed off the stop alone, Dca without the flag, Pulse counting != BRACKETED, the heading without the id, unattributed groups first, grouping by name, the client filtering by rule_name, the client heading taken from the key, the stop cell showing the sentinel, the stop column drawn on the sentinel, the no-stop cell set to warn, and the CLI wording.
  • For each mutant, the script checked that the source had actually changed, then restored it from a saved copy byte for byte.
• keel status was run against a scratch confirm-mode copy of the paper config and DB. The DCA lines read (no stop by design). The PAXG-style turtle tranche (with a stop) and the SOL turtle tranche (without one) read (NO bracket). In paper mode the output was the same as before.
• Real browser: headless Chromium, with the service worker unregistered and caches cleared, against my own keel serve on port 8811 over that scratch copy.
  • Holdings drew six tables, each heading unique, each id free of whitespace: rule-21, rule-101, rule-102, rule-103, unattributed-1 (dca), unattributed-2 (turtle).
  • The DCA cells read "no stop by design" with no warn class. SOL and PAXG read <span class="warn">NO bracket</span>.
  • Pulse showed unbracketed_count 2 (warn), with 4 DCA tranches neutral.
  • There were 0 console errors and no buttons in the tranche tables.

🤖 Generated with Claude Code

[eaitbrahim]

Gatekeeper: READY for the owner to merge (head c686c35)

I ran 5 review/fix rounds with a fresh reviewer each round. Findings per round: 2, 2, 1, 1, 0. All of them are fixed, and none is open.

Changes in the review loop

• Paper is unchanged. A paper tranche's Holdings stop cell and stop list show the stop the bracket verdict names: the raw value in paper, and recorded_stop in live. One helper decides this: status.shown_stop.
• The stop column is decided per table. Each group carries its own stop_rows_present, so the DCA tables draw no stop column while the PAXG table does. Done-means: "stop only when one exists".
• New pins:
  • a negative-stop verdict case;
  • an attributed turtle's stop listed under its own group;
  • per-group stop lists across groups;
  • every registry kind is named after its kind, and no name is stored in its params;
  • keel rules add refuses a name param and writes no row.
• Runbook, docstring and comment wording now says only what is true in both live and paper.

Safety

• "NO bracket" (WARN) still shows for:
  • any live tranche with a real stop (>0) and no bracket, of any kind;
  • every non-DCA tranche without a stop, including PAXG (doctor: nothing warns when an open tranche has no live rule and no retry record -- PAXG sat unwatched for five cycles #811);
  • an unknown or empty kind.
• keel status, the Pulse roll-call and count, and Holdings all read one bracket_verdict.
• Comparing base against head, the live "bracketed" and "NO bracket" output is byte-identical except for the intended DCA change, and paper output is byte-identical.

Evidence on c686c35

• CI: 4/4 green.
• pytest: 8307 passed, 3 skipped. ruff, format and mypy are clean. node --check passes on every static JS file and sw.js.
• Browser (own serve, service worker unregistered):
  • Holdings shows 4 tables, one per rule row, with the unattributed table last.
  • Only the PAXG table has a stop column.
  • The DCA tranches read "no stop by design" (neutral); SOL and PAXG read "NO bracket" (warn).
  • Headers and cells pair up, ids are unique, there are 0 buttons and 0 errors.
  • Pulse shows 2 brackets missing.
• keel status in confirm mode agrees with the browser, and paper strings are unchanged.

Filed outside this PR: #973 (keel positions measures DCA stop distance against the 0 sentinel), #974 (scripts/rule_manifest.py apply can store a name param).
Held: doctor's position.unprotected judges brackets on its own (it agrees on DCA); rule_name is still declared sortable; grouping cost is quadratic, which is fine at today's ~10 tranches.

🤖 Generated with Claude Code