Skip to content

Actions: Kav-K/agent-challenge

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
37 workflow runs
37 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
v1.3.0 Publish to PyPI and npm #6: Release v1.3.0 published by Kav-K
1m 42s
1m 42s
v1.3.0 — Polished chained_arithmetic, expanded template variety Tests #31: Commit 5bc88f6 pushed by Kav-K
1m 23s main
main
1m 23s
Remove knowledge facts from chained_arithmetic — GPT-5.2 drops to 90% Tests #30: Commit 7cff50a pushed by Kav-K
1m 18s main
main
1m 18s
Enhance chained_arithmetic with world-knowledge patterns Tests #29: Commit bc8bb00 pushed by Kav-K
1m 15s main
main
1m 15s
Add knowledge_math type (hard tier) — world-knowledge + arithmetic Tests #28: Commit d895229 pushed by Kav-K
1m 24s main
main
1m 24s
v1.2.0 - Stable release Publish to PyPI and npm #5: Release v1.2.0 published by Kav-K
37s
37s
v1.2.0 — Empirical tier restructure, human-proof agentic challenges Tests #27: Commit c1bd5e3 pushed by Kav-K
1m 13s main
main
1m 13s
Agentic: remove human-trivial types, expand chained_arithmetic Tests #26: Commit d9758aa pushed by Kav-K
1m 25s main
main
1m 25s
Add chained_arithmetic (agentic) + power_mod (hard) types Tests #25: Commit f9dae3f pushed by Kav-K
1m 23s main
main
1m 23s
Relax CI thresholds for new tier structure Tests #24: Commit 81f6913 pushed by Kav-K
1m 4s main
main
1m 4s
Fix CI: dynamic mode test uses gpt-4o solver Tests #23: Commit 6e718a8 pushed by Kav-K
1m 3s main
main
1m 3s
v1.1.0: Tier restructure — GPT-5.2 100% on all tiers Tests #22: Commit a25305a pushed by Kav-K
1m 20s main
main
1m 20s
Dynamic mode: iterated prompt for 100% GPT-5.2 solve rate Tests #21: Commit de41606 pushed by Kav-K
1m 24s main
main
1m 24s
GPT-5.2 calibration data + frontend updates Tests #20: Commit 2972ae4 pushed by Kav-K
1m 26s main
main
1m 26s
Add 14 safe_solve exact-answer enforcement tests (217 total) Tests #19: Commit 78cfd4b pushed by Kav-K
1m 23s main
main
1m 23s
v1.1.0 — Empirical tier reclassification, agentic dynamic mode, safe_… Tests #18: Commit ba07d08 pushed by Kav-K
1m 26s main
main
1m 26s
safe_solve() exact-answer enforcement + calibration data Tests #17: Commit af089a3 pushed by Kav-K
1m 15s main
main
1m 15s
Make wrong-answer retry test resilient (3 retries) Tests #16: Commit 9a79d5d pushed by Kav-K
1m 31s main
main
1m 31s
Lower medium bulk threshold to 75% for CI variance Tests #15: Commit 1053fd8 pushed by Kav-K
1m 30s main
main
1m 30s
Empirical difficulty reclassification with dual-model validation Tests #14: Commit e33bc37 pushed by Kav-K
1m 26s main
main
1m 26s
Adjust easy threshold to 80% (16/20) for CI variance Tests #13: Commit e468e8a pushed by Kav-K
1m 19s main
main
1m 19s
Reclassify difficulty tiers based on empirical gpt-4o-mini accuracy Tests #12: Commit c174193 pushed by Kav-K
1m 23s main
main
1m 23s
v1.0.0 - Initial release Publish to PyPI and npm #4: Release v1.0.0 published by Kav-K
35s
35s
v1.0.0 — Production release Tests #11: Commit 1dd1ff5 pushed by Kav-K
1m 15s main
main
1m 15s
Fix live test timeout: increase to 45s, wrap bulk solves in try/except Tests #10: Commit edf8b07 pushed by Kav-K
1m 11s main
main
1m 11s