SkillTrialNew run
Reports

Published comparisons

Anchored on Solana by their authors. Open any of them to see every trial, the reasons and the raw answers

solana-token-helper anchored
1.4.2 → 1.5.0 · Claude Sonnet 5.5 · 3 reps

v1.5.0 vs v1.4.2: better on TX-03, worse on PLAN-05

base 9/18A 15/18B 16/18
r_7Hk2QpV9 ·
spl-explainer anchored
2.1.0 → 2.2.0 · Claude Sonnet 5.5 · 3 reps

v2.2.0 vs v2.1.0: better on TX-03

base 8/18A 12/18B 14/18
r_3mXa81Ld ·
tx-reader anchored
0.9.3 → 1.0.0 · Claude Haiku 5.5 · 3 reps

v1.0.0 vs v0.9.3: no observed difference

base 8/18A 12/18B 12/18
r_Qe5tYw2n ·
anchor-idl-guide anchored
3.0.1 → 3.1.0 · Claude Sonnet 5.5 · 5 reps

v3.1.0 vs v3.0.1: worse on TX-04

base 9/30A 22/30B 20/30
r_Z8cJr4Fs ·
token22-notes anchored
1.2.0 → 1.3.0-rc1 · Claude Sonnet 5.5 · 3 reps

Incomparable — no automatic winner

base 7/18A 13/18B 10/18
r_9vNb6KpE ·
wallet-plan-kit anchored
0.4.0 → 0.5.0 · Claude Haiku 5.5 · 3 reps

v0.5.0 vs v0.4.0: better on TX-03, TX-04

base 4/18A 9/18B 13/18
r_Lw4Ds0Mh ·

Showing 6-task solana-core reports. Reports are only comparable at the same pack, checker and model revision.