Raw totals are only compared when every candidate computed the same
(event, layer) evidence units. Otherwise winning_segment stays the raw
leader (provisional_leader) and comparable_leader re-ranks over the
units all candidates share; a missing unit is dropped for everyone,
never scored 0. The live stage Python engine now skips an event whose
Narayana is unavailable instead of scoring it 0, repeated facts count
once, and the stage labels an unequal-coverage leader as provisional.
占星师 2026-11-08 验收清单 §6 / §9 / §10.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
- R1: default AD direction rath_p50_md_parity_v1 (start parity kept as a research
profile); full profile config + algorithm version on every Narayana output, cache
keys (consult, api chart, probes, long report, synastry) and the rectification run.
- R2: one strength profile at the seed, AD start and p.43 (d); selected / equal /
unresolved kept apart; unresolved nodes block only what depends on them (seed
candidates kept, dated MDs before an unresolved length kept, AD-only loss in
rectification keeps the MD).
- R3: Neecha Bhanga witness fields (reference, relative_house, self_reference,
independent_confirmation_allowed, self_reference_only) from the main detector to
the evidence card yogas_detail rows; self-reference-only never counts as support.
- Sync her b54812c3 KP clock validation; port her round-5 / round-6 tests.
- BUG-1783..1787; golden recaptured with TTL 0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
- seed rath_highest_occupant_dignity_v1 (named convention; equal tier below own sign blocks the seed)
- AD direction profiles rath_p50_md_parity_v1 / rath_stronger_lord_start_parity_v1 (default, pending choice)
- Saturn+Ketu in seed / MD sign blocked; PyJHora Ketu priority only as named reference profile
- 0-year MD rows flagged, never current, skipped on the annual timeline; years profile rath_table8_phalita_dignity_v1
- KP DBA year_days 365.25 default, parent/AD/full/visible fields, boundaries equal main Vimshottari
- functional profile pvr_table30_v1 (aliases), Moon-only blocked without paksha; profile through API,
cache keys, audit; rectification asserts one profile per run
- golden recaptured with TTL=0
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
Product approved 2026-11-08 (astrologer round-5 rulings 1, 3; same basis as
the report and the chat card):
- functional roles: drop the bphs_exception_profile pin at the three scoring
call sites (+ research mirror); default pvr_table_30 with Moon paksha.
- Narayana: narayana_rath.calc_narayana_rath (two cycles) / antardasha, ages
via narayana_age from the candidate birth moment; probes date from the birth
moment. A blocked source step makes the layer unavailable, never the old
algorithm. Blocked count on all three datasets: 0.
- Reseal: sealed 20, offset 900, live v5 231 under the new tag; identical to
measurement variant (c) trial by trial. float_split / 166d kept.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
- Functional default = PVR Table 30 (functional_profile=pvr_table_30); the former
product table is bphs_exception_profile, scored only when named (rectification
names it, scores unchanged). profile_divergence shows both roles side by side;
movable-lagna Moon by the table's paksha note; dual-lordship note; prior rulings
BUG-1570/1590/#166 Q4 checked against Table 30.
- Shadbala: six open items labelled parameter_sensitive; every option combination
computed; only planets whose pass/fail flips get minimum_check=parameter_sensitive.
- Vimshottari AD/PD rows carry parent_period_start / visible_period_start; the
consult card cuts the balance MD from its full parent and clips at birth.
- KP DBA: full first parent, clipped at birth (no balance rescale).
- Mudda: balance fields + balance_method (angle default, time comparison apart),
nominal_year_days, her crossing wording; Harsha combined blocked reason lists items.
- D30 _meta describes the unequal parts (as her r40).
- Golden recaptured with TTL=0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
Product approved 2026-11-08. _active_vimshottari, the candidate context
timeline, chart_choice astrologer and every event_probes Vimshottari start
now use birth_moment.mahadashas from the candidate birth moment and
birth_moment.split_period for AD/PD, replacing build_dasha_timeline
(00:00 of the birth date) and the integer-day build_antardasha. Frozen
dasha_analyzer.py / domain_calculation_service.py unchanged. Research
mirrors follow. A bare date string now raises TypeError.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyn7Hj3LCW3BUs5bZ8
The legacy current-period lookup took the age modulo the first-cycle total,
so anyone older than their first cycle (61-96 years on the nine public
charts) got a first-cycle period: an 80-year-old was placed in 2-12. The
second cycle runs the same signs for 12 minus their first-cycle years; the
lookup now places the age on both cycles and subdivides with the second
cycle's years. The rectification cache only covers the first cycle and falls
back to the live lookup past it, so first-cycle scoring is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbBXoyeZrYn6ryNcSe5Y2Y
Cache work that depends only on the candidate minute, drop unread response fields, and reject windows above 241 candidates before any chart is built. Eleven canonical scenario hashes match 77057a819. Sealed and reported-offset replays match the 2026-10-08 records (changed_trial_count 8 and 581) and are stored as scoring_perf_2026_10_13.
Co-Authored-By: Grok (agent loop)
Carry explicit local date intervals instead of inferring the day from clock
order. Cluster width, delivery, adoption, and reports keep the actual civil
date; adopted date is stored separately from the reported birth_date.
Algorithm identity is scoring-9 / spec-v5. Scoring weights, confirmation
thresholds, and Skill version are unchanged. Isolated Linux final-3 gates
passed; four pre-existing Python failures remain. This is not a production
release.
Keep askable cards after exhaustion, explain each probe, read the adopted credible range in reports and chat, and compare declared periods before the minute grid when the clock is unknown.
Co-authored-by: Cursor <cursoragent@cursor.com>
Interviews skipped D2/D11 and D30 unless the user volunteered them, so those scoring layers almost never ran. Ask them after family, keep relocation out of rotation, and display Bhava/Pranapada transitions without delaying adopt.
Co-authored-by: Cursor <cursoragent@cursor.com>
Window scan now drives d5_refine when D24 changes, family events use D3, and pada/Hora/Ghati are display-only. New cases bind Skill 10.0.7 without unique-minute confirmation.
Co-authored-by: Cursor <cursoragent@cursor.com>
Keep family on D12+D7 instead of mixing it into D4, score D5 on education, and show a natal recast plus technique audit after adopt without unique-minute claims.
Co-authored-by: Cursor <cursoragent@cursor.com>
Dated family evidence now moves candidates via D12 and kin houses, career receipts expose both D1-10 and D10, and appearance/marks may be asked as auxiliary first-house scores. New cases bind Skill 10.0.4.
Co-authored-by: Cursor <cursoragent@cursor.com>
ClaudeandClaude Opus 5.5