Skip to content

Actions: Balisa50/ayat

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
23 workflow runs
23 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
polish: tween layout, anchor top, header chevrons CI #23: Commit ae07902 pushed by Balisa50
1m 9s main
main
1m 9s
fix: middle-dot rendering, invisible edge tap zones CI #22: Commit b1edff5 pushed by Balisa50
1m 6s main
main
1m 6s
fix: in-surah nav, tap zones, audio continuity, smooth layout CI #21: Commit 8c212da pushed by Balisa50
1m 8s main
main
1m 8s
fix: restore working VerseCard CI #20: Commit 38778b5 pushed by Balisa50
1m 2s main
main
1m 2s
fix: restore working VerseCard CI #19: Commit 5361f4e pushed by Balisa50
1m 4s main
main
1m 4s
fix: history uses functional updater, header stacks on mobile CI #18: Commit 19000d5 pushed by Balisa50
39s main
main
39s
fix: history persists, mobile Arabic spacing, smooth card transitions CI #17: Commit 043b2c1 pushed by Balisa50
1m 2s main
main
1m 2s
fix: keep reading history when card closes CI #16: Commit 3b40435 pushed by Balisa50
1m 6s main
main
1m 6s
fix: always show nav strip, add Arabic and transliteration retrieval CI #15: Commit 36ea833 pushed by Balisa50
1m 1s main
main
1m 1s
fix: ask refuses to invent verses, adds explicit-ref and surah-name c… CI #14: Commit 35891b8 pushed by Balisa50
1m 4s main
main
1m 4s
fix: NVIDIA uses nemotron-70b, Groq uses qwen3.8-27b CI #13: Commit d7dab24 pushed by Balisa50
1m 1s main
main
1m 1s
fix: switch to non-reasoning models, bump maxTokens for full responses CI #12: Commit c6fab04 pushed by Balisa50
1m 4s main
main
1m 4s
fix: restore the retry in the ask route, fall through to text search … CI #11: Commit 2d20087 pushed by Balisa50
53s main
main
53s
fix: search falls back to text matches when AI is slow or down CI #10: Commit 8ff6240 pushed by Balisa50
1m 11s main
main
1m 11s
perf: align reflect route timeouts with context route CI #9: Commit b3388e2 pushed by Balisa50
1m 2s main
main
1m 2s
fix: switch Groq to gpt-oss-120b CI #8: Commit a868770 pushed by Balisa50
1m 6s main
main
1m 6s
perf: shorter timeouts on tafsir calls, one attempt per model CI #7: Commit 31e9861 pushed by Balisa50
1m 1s main
main
1m 1s
fix: drop retired gpt-oss-120b fallback, give the reasoning model roo… CI #6: Commit 4d90aba pushed by Balisa50
1m 0s main
main
1m 0s
fix: use openai/gpt-oss models on NVIDIA NIM after 2026-08 retirements CI #5: Commit c3793bd pushed by Balisa50
1m 7s main
main
1m 7s
ops: health check can run a live completion to surface the real error CI #4: Commit d32a452 pushed by Balisa50
1m 8s main
main
1m 8s
ops: add /api/health/ai so dead provider keys are visible CI #3: Commit 18642ca pushed by Balisa50
1m 9s main
main
1m 9s
docs: drop the em dashes and ticks from the README CI #2: Commit d53dac5 pushed by Balisa50
1m 6s main
main
1m 6s
ci: make lint work again, and put the failover tests in CI CI #1: Commit 622b840 pushed by Balisa50
1m 7s main
main
1m 7s