root/Jyotisha
1
0
Fork 0
Code Issues 96 Pull Requests Actions Packages Projects Releases Wiki Activity

建档：按出生时间范围分项算稳定性（#144 第二单） #165

New Issue
Open
opened 2026-11-02 17:46:04 +08:00 by Claude · 3 comments
Claude commented 2026-11-02 17:46:04 +08:00
Collaborator
Copy Link
Copy Source

#144 第二单（jesse 11-02 选 A：拆两单）。前置：#144 第一单合入 staging。

要做什么

  • 上升预检从「只看 D1 上升星座变不变」扩成按用户的时间范围分项实际计算：
    • 上升星座；上升度数和宫头（分开说，星座不变不代表度数宫头不变）；
    • D9、D10 的上升（接口已有，现在没请求）；
    • 月亮的星座、星宿、pada、宫位（四项分开，现在都没算）。
  • 每项只说「在您给的范围内一致 / 有变化」，必须来自计算，不预先写死；不把某个分盘上升变了扩大成整张分盘无效。
  • 范围比较完成后才显示原文：「我们已比较您提供的时间范围，并标出了保持一致和发生变化的结果。」
  • 大运：不再写「准到几天」「前后两到四个月」之类固定承诺；用原文：「出生时间的不确定性可能影响周期边界。我们会说明计算采用的时间和方法；计算出的日期不代表事件一定发生的日期。」
  • 统一计算设置：上升预检、candidate_time_sensitivity_scan.py、flexible_birth_time_profile.py 现在岁差默认 Raman、交点写死平均交点；改成每次显式传入档案上的岁差和交点设置。
  • 何时先核实时间（不做成所有新用户的门槛）：不同来源时间冲突、日期 / 上下午 / 跨午夜有疑问、时区或历史夏令时没确认、问题依赖的结果在范围内实质变化 → 提示先核实原始资料，必要时再进生时校正；校正说明用原文「生时校正会比较不同候选时间与已知经历的吻合情况，帮助缩小范围；它不能保证找回真实的出生分钟。」
  • 只记得时辰的第三个问题：「出生日期确定吗？如果接近午夜，家里对是哪一天有没有不同说法？」

文案全部用 #144 jesse 11-02 10:54 评论的原文。第一单合入后由循环写任务书。

#144 第二单（jesse 11-02 选 A：拆两单）。**前置：#144 第一单合入 staging。** ## 要做什么 - 上升预检从「只看 D1 上升星座变不变」扩成**按用户的时间范围分项实际计算**： - 上升星座；上升度数和宫头（分开说，星座不变不代表度数宫头不变）； - D9、D10 的上升（接口已有，现在没请求）； - 月亮的星座、星宿、pada、宫位（四项分开，现在都没算）。 - 每项只说「在您给的范围内一致 / 有变化」，**必须来自计算，不预先写死**；不把某个分盘上升变了扩大成整张分盘无效。 - 范围比较完成后才显示原文：「我们已比较您提供的时间范围，并标出了保持一致和发生变化的结果。」 - 大运：不再写「准到几天」「前后两到四个月」之类固定承诺；用原文：「出生时间的不确定性可能影响周期边界。我们会说明计算采用的时间和方法；计算出的日期不代表事件一定发生的日期。」 - **统一计算设置**：上升预检、`candidate_time_sensitivity_scan.py`、`flexible_birth_time_profile.py` 现在岁差默认 Raman、交点写死平均交点；改成每次显式传入档案上的岁差和交点设置。 - 何时先核实时间（不做成所有新用户的门槛）：不同来源时间冲突、日期 / 上下午 / 跨午夜有疑问、时区或历史夏令时没确认、问题依赖的结果在范围内实质变化 → 提示先核实原始资料，必要时再进生时校正；校正说明用原文「生时校正会比较不同候选时间与已知经历的吻合情况，帮助缩小范围；它不能保证找回真实的出生分钟。」 - 只记得时辰的第三个问题：「出生日期确定吗？如果接近午夜，家里对是哪一天有没有不同说法？」 文案全部用 #144 jesse 11-02 10:54 评论的原文。第一单合入后由循环写任务书。
Claude added the agent/试点agent/排队 labels 2026-11-02 17:46:07 +08:00
Claude commented 2026-11-02 22:36:27 +08:00
Author
Collaborator
Copy Link
Copy Source

这一单接在 #144 后面，也改由同一个 Claude 会话直接做（jesse 已定），不再排队等 Grok。#144 上了测试环境之后接着做。

<!-- agent:claude plan executor=claude-session handover=jyotisha-9c --> 这一单接在 #144 后面，也改由同一个 Claude 会话直接做（jesse 已定），不再排队等 Grok。#144 上了测试环境之后接着做。
Claude removed the agent/排队 label 2026-11-02 22:36:30 +08:00
Claude commented 2026-11-03 11:16:02 +08:00
Author
Collaborator
Copy Link
Copy Source

#165 on staging (fast-forward 5213a861 → ae1e9ada; no migration)

  • CI: backend-quality-gate 6455 ✅ · migrate-staging-database 6456 ✅ (nothing new to apply) · deploy-staging 6457 ✅
  • Smoke on https://staging.jyotisha.chat: /login 200 · /api/account 401 · /api/health 200 · /chart 200 · /learn 200

What shipped:

  • Per-item stability over the user's range. These items are computed, not hard-coded:
    • lagna sign
    • lagna degree
    • house cusps
    • D9 lagna
    • D10 lagna
    • Moon sign, nakshatra, pada and house
  • Comparison and dasha copy, verbatim.
  • The verify-first prompt appears only when a condition is detected. It is not a gate for new users.
  • The precheck, candidate_time_sensitivity_scan.py and flexible_birth_time_profile.py now take ayanamsa and node settings explicitly.
  • The third 时辰 question, stored in birth_time_clue.
  • BUG-1515: a scan without an ayanamsa computed in Raman but fingerprinted Lahiri.

Needs product confirmation:

  • "Material change" on the chart page means a change in the D1 lagna sign or the Moon nakshatra.
  • The labels for the third question (确定 / 有不同说法 / 不确定) and for the nine items were written by us. They are not verbatim from the issue.
  • 「准到几天」 / 「两到四个月」 no longer exist anywhere in the repo, so there was nothing to replace.

Known gaps:

  • No question-domain verify-first in consultations or reports.
  • The precheck uses a single timezone offset across the whole range. A DST change is flagged but not recomputed.
  • Conflicting sources and AM/PM doubt aren't stored, so they aren't detected.

Device checklist: docs/testing/birth-time-stability-20261103.md

<!-- agent:claude deployed branch=codex/birth-time-stability-165 sha=ae1e9ada42e9ce019ff46d8b29b4c9371568fd96 --> **#165 on staging** (fast-forward 5213a861 → ae1e9ada; no migration) - CI: backend-quality-gate 6455 ✅ · migrate-staging-database 6456 ✅ (nothing new to apply) · deploy-staging 6457 ✅ - Smoke on https://staging.jyotisha.chat: /login 200 · /api/account 401 · /api/health 200 · /chart 200 · /learn 200 What shipped: - Per-item stability over the user's range. These items are computed, not hard-coded: - lagna sign - lagna degree - house cusps - D9 lagna - D10 lagna - Moon sign, nakshatra, pada and house - Comparison and dasha copy, verbatim. - The verify-first prompt appears only when a condition is detected. It is not a gate for new users. - The precheck, `candidate_time_sensitivity_scan.py` and `flexible_birth_time_profile.py` now take ayanamsa and node settings explicitly. - The third 时辰 question, stored in `birth_time_clue`. - BUG-1515: a scan without an ayanamsa computed in Raman but fingerprinted Lahiri. Needs product confirmation: - "Material change" on the chart page means a change in the D1 lagna sign or the Moon nakshatra. - The labels for the third question (确定 / 有不同说法 / 不确定) and for the nine items were written by us. They are not verbatim from the issue. - 「准到几天」 / 「两到四个月」 no longer exist anywhere in the repo, so there was nothing to replace. Known gaps: - No question-domain verify-first in consultations or reports. - The precheck uses a single timezone offset across the whole range. A DST change is flagged but not recomputed. - Conflicting sources and AM/PM doubt aren't stored, so they aren't detected. Device checklist: `docs/testing/birth-time-stability-20261103.md`
Claude added the agent/待真机 label 2026-11-03 11:16:03 +08:00
Claude commented 2026-11-03 11:18:33 +08:00
Author
Collaborator
Copy Link
Copy Source

Product decision (jesse, 2026-11-03): all three agreed as shipped

  1. On the chart page, a "material change" (which triggers the verify-first prompt) means the D1 lagna sign or the Moon nakshatra changes within the range. A change in D9/D10 lagna or Moon house alone does not trigger it.
  2. The third-question options 确定 / 有不同说法 / 不确定 and the nine item labels stay as written.
  3. 「准到几天」 / 「两到四个月」 don't exist in the repo, so no replacement is needed.

No further code change. Remaining: device check (docs/testing/birth-time-stability-20261103.md).

<!-- agent:claude decision --> **Product decision (jesse, 2026-11-03): all three agreed as shipped** 1. On the chart page, a "material change" (which triggers the verify-first prompt) means the D1 lagna sign or the Moon nakshatra changes within the range. A change in D9/D10 lagna or Moon house alone does not trigger it. 2. The third-question options 确定 / 有不同说法 / 不确定 and the nine item labels stay as written. 3. 「准到几天」 / 「两到四个月」 don't exist in the repo, so no replacement is needed. No further code change. Remaining: device check (`docs/testing/birth-time-stability-20261103.md`).
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
agent/待执行
Grok 领：按任务书实现
 agent/待真机
产品负责人真机走查
 agent/待规划
Claude 领：诊断并写任务书
 agent/待验收
Claude 领：验收 Grok 回报的 SHA
 agent/执行中
锁：Grok 执行中
 agent/排队
任务书已写好，等前置单合入 staging 后由 Claude 改成 agent/待执行；循环两侧都不处理
 agent/规划中
锁：Claude 规划中
 agent/试点
试点期只处理带此标签的 issue
 agent/部署中
已推 staging，等部署核对（脚本自动）
 agent/需拍板
需要产品负责人决定
 agent/验收中
锁：Claude 验收中
 ready-for-agent
Fully specified, ready for an agent
 wayfinder:map
Wayfinder map issue holding the notes and child tickets
 wayfinder:task
Wayfinder child ticket
No labels agent/待真机 agent/试点
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
Claude Grok jesse root
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: root/Jyotisha#165