Work Tracker

✍️ Day Summary

I corrected the Rahu Kaal and other time periods within the muhurta calculations. I also developed a new summary endpoint for muhurta information and updated the system to track changes in the muhurta API. Finally, I documented these changes regarding muhurta segregation and input handling.


📊 Daily Overview

Total Activity: 5 commits across 1 repositories


🚀 GitHub Activity

1. astroacharya 🔒 — 5 commits

fix(sky_sun): start rise/set search at local midnight, not 0h UT · 40c6462

swe.rise_trans started its search at swe.julday(y,m,d,0.0) = 0h UT instead
of local midnight. For east-of-Greenwich locations whose sunrise clock is
earlier than the tz offset (Delhi summer ~05:25 IST = 23:55 UT the prior
day; any high-tz Pacific location), the search skipped that day’s sunrise
and returned the NEXT day’s, with a near-identical clock value.

The error was clock-invisible — every HH:MM comparison (incl. drikpanchang
cross-checks) passed — but the raw JD that yearly_calendar.py feeds into
sun/moon longitude sampling was ~24h off on affected days, shifting
tithi/nakshatra by up to one unit.

fix(muhurta): correct Rahu Kaal, Gulika/upagraha sequence, add Yamaganda + day boundary · 88d37ba

Verified against drikpanchang (New Delhi Thu/Sun, Mumbai Mon).

  • Rahu Kaal: _RAHU_KAAL_INDEX was uniformly one part too early on every
    weekday. Corrected to the standard table (Sun=8, Mon=2, Tue=7, Wed=5,
    Thu=6, Fri=4, Sat=3); the 1st (sunrise) part is never Rahu’s.
  • Gulika/Kala/Mrityu/Yamaghantaka/Ardhaprahara: the 8-part lord tables were
    generated from the Chaldean HORA sequence instead of the VARA (weekday)
    sequence that BPHS and drikpanchang use.

docs(plans): muhurta segregation + input-hardening design spec · 3f7828a

feat(muhurta): add /muhurta/summary tier + shared input-hardening dependency · 8ad47b8

Segregates the muhurta surface into a small summary tier and the existing
detailed API, and fixes lat/lon/tz handling once in a shared dependency.

Summary tier — GET /muhurta/summary:

  • 5 headline windows (Rahu Kaal, Yamaganda, Gulika, Abhijit, Brahma) + their
    bilingual descriptions + a now-aware current/next pointer.
  • A projection of the SAME compute_day_segment_windows the detailed API uses
    (projection invariant tested: summary windows == detailed windows), so the
    tiers can never disagree.
  • Lean prelude: computes only sunrise/sunset (one rise_trans pair) + vara
    (pure calendar math), NOT full get_panchanga (which computes
    tithi/nakshatra/yoga/karana the card never shows).

chore(propagate): track muhurta API → consumers in sidecar · 9252d84

Declares app/api/muhurta.py as a propagation source so endpoint/param/
response changes fire ‘verify consumers’ drift. Downstreams: VK’s
astroacharya client, muhurta service wrapper, and MuhurtaSection. Astroclarity
omitted pending rebuild (matches existing convention); its integration is
tracked in Astroclarity/STATE.md. Ships with the /muhurta/summary segregation.


📈 Development Analytics

Daily Summary

MetricToday
Commits5
Merges0
Skill runs0
Decisions0
Eurekas0
CEO plans0
Notes touched0

Decisions logged today

None today.

Eurekas

None today.

Strategic plans authored

None today.

Notes touched

None today.

Running totals (as of this date)

Active days last 90:           69 / 90  (substantive)
Days with any signal last 90:  72 / 90
Skill runs (since 2026-03-25):  75
Commits this year (authored):  1439
Decisions logged this year:    15
Eurekas this year:             5
CEO plans authored this year:  13
Projects touched this year:    57
Top projects last 90:          VipinKaushik (27d), astroacharya (20d), portfolio (20d), vipin-kaushik-content (17d), Mothership (13d)

💭 My Notes

[Your own thoughts, observations, or context for this day — this section is never overwritten by automation]