root/Jyotisha
1
0
Fork 0
Code Issues 1 Pull Requests 1 Actions Packages Projects Releases Wiki Activity
Files
bedb4d7bd12d02acea843b260dba1dc8399abfea
Jyotisha/docs/research/ACTIVE_FRONTS.md
T
jesse-uxandClaude Code bedb4d7bd1 research(rectification): archive partial varga-resolution study (BUG-1105) 
Archive research scripts, regression tests, M1 results and safe M0 smoke. Keep the incomplete study and failing quick gate explicit. Exclude full M0 JSON, raw logs and unrelated oracle newline changes.

Co-Authored-By: Claude Code <noreply@anthropic.com>
2026-09-30 08:49:54 +08:00

8.5 KiB
Raw Blame History

Active Fronts

This file is the small index for the current engineering fronts that still drive code changes.

生时校正验证完整性（2026-09-20）

  • 申报偏差敏感性：申报中心搜索窗与真值出窗，不能当真实用户准确率。
  • 固定口径重跑与契约：已曝光 v3 的当前实现成绩，不是新的独立盲测；确认门关闭。
  • v5 独立封存采集协议：本轮只出协议，不采事件。
  • 前置结论：分钟分辨率两轮研究。不重走已证伪的加权重与簇合并假设。
  • 验证补缺轮：进度。
  • 后续候选级跨午夜日期修复：进度与独立验收。当前评测转至 reported_offset_cross_midnight_2026_09_20.json / sealed_holdout_rerun_cross_midnight_2026_09_20.json；旧 JSON / freeze 原字节留存，旧文档及评测器在 history/rectification_pre_cross_midnight_2026_09_20/。
  • 尚未闭环：BUG-982 跨午夜簇跨度、BUG-983 凌晨日期锚点、BUG-984 时段缓存/回执身份；验收补单 不代表批准实施。

Shortest-Path Closure Order (2026-06-29)

To avoid scope drift, current work stays inside these four closure lanes only:

  1. Relationship adjudicator closure
  2. Vimsopaka + functional-role closure
  3. Oracle closure batch
  4. VedAstro strict ingestion

Reference plan:

  • <repo>/docs/superpowers/plans/2026-06-29-shortest-path-closure-plan.md

Do not open new product surfaces before at least one of these four lanes is closed.

Fragment Discipline

  • Before touching strict workflow or adjudicator logic, check:
    • <repo>/docs/research/high_value_fragment_source_map_2026_06_28.md
    • <repo>/docs/research/worktree_fragment_triage_2026_06_28.md
    • <repo>/docs/research/yoga_benchmark_case_promotion_audit_2026_06_28.md
    • <repo>/docs/research/promote_first_repo_truth_pack_2026_07_01.md
    • <repo>/docs/research/promote_second_repo_truth_pack_2026_07_01.md
    • <repo>/docs/research/promote_third_repo_truth_pack_2026_07_01.md
    • <repo>/docs/research/promote_fourth_repo_truth_pack_2026_07_01.md
  • Treat repo truth as authoritative.
  • Treat docs/research/local_drafts, Gemini brain notes, Codex attachments, and WorkBuddy copies as candidate sources only.

Relationship Adjudication

  • <repo>/docs/research/marriage_adjudicator_first_pass_audit_2026_06_27.md
  • <repo>/docs/research/marriage_benchmark_summary_bridge_audit_2026_06_28.md
  • <repo>/docs/research/isolated_asset_bridge_audit_2026_06_28.md
  • <repo>/references/event_judgment_marriage.md
  • <repo>/docs/superpowers/specs/2026-06-28-jaimini-marriage-bridge-v1-design.md
  • <repo>/scripts/marriage_benchmark_summary.py

Career Adjudication

  • <repo>/references/event_judgment_career.md
  • <repo>/tests/test_mcp_strict_workflow_career.py
  • <repo>/docs/research/life_event_graph_v1_audit_2026_06_28.md
  • <repo>/docs/research/functional_benefic_malefic_strict_layer_audit_2026_06_28.md

Wealth Adjudication

  • <repo>/docs/research/wealth_adjudicator_first_pass_audit_2026_06_27.md
  • <repo>/docs/research/wealth_adjudicator_second_pass_audit_2026_06_27.md
  • <repo>/docs/research/wealth_adjudicator_third_pass_audit_2026_06_27.md
  • <repo>/docs/research/wealth_adjudicator_fourth_pass_audit_2026_06_28.md
  • <repo>/docs/research/wealth_adjudicator_fifth_pass_audit_2026_06_28.md
  • <repo>/docs/research/wealth_adjudicator_sixth_pass_audit_2026_06_28.md
  • <repo>/docs/research/wealth_adjudicator_sixth_pass_avayogi_boundary_2026_06_28.md
  • <repo>/docs/research/wealth_adjudicator_seventh_pass_ashtakavarga_bridge_2026_06_28.md
  • <repo>/references/event_judgment_wealth.md
  • <repo>/references/yogi-asc-tight-orb-wealth-freeze-guide.md

Oracle Closure

  • <repo>/docs/research/oracle_benchmark_inventory_latest.md
  • <repo>/docs/research/oracle_benchmark_inventory_latest.json
  • <repo>/docs/research/oracle_closure_master_dashboard_latest.md
  • <repo>/docs/research/public_benchmark_dashboard_latest.md
  • <repo>/docs/research/shadbala_absolute_oracle_comparison_audit_2026_06_28.md
  • <repo>/docs/research/promote_first_repo_truth_pack_2026_07_01.md
  • <repo>/docs/research/promote_second_repo_truth_pack_2026_07_01.md
  • <repo>/docs/research/promote_third_repo_truth_pack_2026_07_01.md
  • <repo>/docs/research/promote_fourth_repo_truth_pack_2026_07_01.md
  • <repo>/docs/research/raman_dasha_boundary_series_oracle_seed_2026_06_28.md
  • <repo>/docs/research/tajika_annual_closure_status_latest.md
  • <repo>/docs/research/tajika_annual_benchmark_dashboard_latest.md
  • <repo>/scripts/shadbala_oracle_comparison.py
  • <repo>/references/oracle/
  • Before changing oracle-dependent adjudicator or benchmark claims, run:
    • python3 scripts/oracle_benchmark_inventory.py --format json

VedAstro Adapter MVP

  • <repo>/docs/research/vedastro_parity_matrix_latest.md
  • <repo>/docs/research/vedastro_parity_matrix_latest.json
  • <repo>/docs/research/vedastro_fast_path_checklist_latest.md
  • <repo>/docs/research/vedastro_fast_path_checklist_latest.json
  • <repo>/docs/research/life_event_graph_v1_audit_2026_06_28.md
  • <repo>/docs/research/vedastro_range_scan_allowlist_audit_2026_06_28.md
  • <repo>/scripts/vedastro_python_bridge.py
  • <repo>/scripts/vedastro_official_mcp_bridge.py
  • <repo>/scripts/vedastro_method_catalog_sync.py
  • <repo>/scripts/vedastro_fast_path_checklist.py
  • <repo>/scripts/vedastro_service_adapter.py
  • follow only after the Jaimini marriage bridge v1 regression loop is closed
  • Use the parity matrix before adding or claiming VedAstro-equivalent capability.
  • Use the fast-path checklist when deciding whether a VedAstro-facing feature should go through official MCP, official Python bridge, REST adapter, or remain local-native.

Dignity / Role Guardrails

  • <repo>/docs/superpowers/specs/2026-06-28-dignity-guardrail-v1-design.md
  • <repo>/docs/research/dignity_guardrail_v1_boundary_audit_2026_06_28.md
  • <repo>/docs/research/divisional_dignity_context_repair_audit_2026_06_28.md
  • <repo>/docs/research/functional_benefic_malefic_strict_layer_audit_2026_06_28.md
  • D1-only dignity guardrail is landed.
  • Divisional dignity context repair for D9/DK/Vimsopaka Navamsa is also landed.
  • Still open:
    • Vimsopaka semantic mapping for NEECHA_BHANGA / GREAT_FRIEND / GREAT_ENEMY
    • functional role now enters strict evidence; follow-up is Technique Audit Table rendering.

生时校正：分盘上升段研究（2026-09-30，BUG-1105）

  • 任务书、进度、阶段性报告。仅离线研究，无生产实现授权。
  • M0 首轮与 M1 全量各 308 个 case-radius 已在本地落盘；本次 staging 归档仅含研究代码/测试、M1 JSON、测试必需的 M0 smoke 和文档，M0 全量原始 JSON 与运行日志因交付检查缺口不提交。M0 起点表逐格一致性、两次字节复跑尚未闭环，不能称整单验收通过。
  • M2/M3 not_started，M4 仅阶段性建议；±60 D9/D10 真值段保留各 74/77，低于既有区间基线 75/77，不通过研究红线。筛选子集的高准确率不能冒充全集或唯一段准确率。

生时校正：候选分不开（2026-09-14 收口）

  • 定论页（先读这个）：<repo>/docs/research/rectification_minute_resolution_closure_2026_09_14.md
  • 两个已被证伪的假设：加权重能拉开候选；簇合并把候选并没了（合并从未触发）。
  • 未解决的只剩「区间内部并列」，目前唯一被证明有效的信息源是用户提供的带年月新经历。
  • 2026-09-29 打分方法研究（脚手架，判定待 v5）：<repo>/docs/research/rectification_scoring_research_2026_09_29.md——似然比校准权重 / 缺席证据 / 精度追问三条流水线已在 v4 上跑通，所有判定 pending_v5；任务书 docs/tasks/TASK-rectification-scoring-research-20260929.md，前置 TASK-rectification-holdout-expansion-20260929.md。
  • 2026-09-29 起评价集换 v5（77 例公开 AA，references/real_case_calibration/minute_rectification_holdout_v5.json）：协议 holdout_v5_protocol_2026_09_29.md、构建与基线成绩单 holdout_v5_build_2026_09_29.md。后续打分判定以 v5 为准，v4 数字只作历史对照。
Reference in New Issue View Git Blame Copy Permalink