Codex CLI 0.104.0

GPT 5.2 2026-02-24
Score
19.01
PASSED
Pass Rate: 61.5%
go: 83.3% (6 tasks)
go
typescript: 80.0% (5 tasks)
typ
kotlin: 66.7% (3 tasks)
kot
rust: 50.0% (6 tasks)
rus
dart: 33.3% (3 tasks)
dar
zig: 33.3% (3 tasks)
zig

Evaluation Report

Summary

Metric Value
Agent codex
Model gpt-5.2
Reasoning Effort medium
Legacy Mode Yes
Timestamp 2026-02-24T150330
Pass Rate 61.5% (16/26)
Weighted Pass Rate 57.1%
Weighted Score 19.01 / 33.29
Duration 3785.1s

Quality Breakdown

  • Clean Passes: 15
  • Partial Passes (timed out but passed): 1
  • Integrity Violations (modified test files): 0
  • Failures: 9

Results by Language

Language Passed Failed Total Pass Rate
dart 1 2 3 33.3%
go 5 1 6 83.3%
kotlin 2 1 3 66.7%
rust 3 3 6 50.0%
typescript 4 1 5 80.0%
zig 1 2 3 33.3%

Results by Tier

Tier Passed Failed Total Pass Rate
core 8 4 12 66.7%
extended 8 6 14 57.1%

Task Results

Task Status Weight Score Duration
dart/future-pool ✘ FAIL 1.46 0.00 162.6s
dart/isolate-pool ✘ FAIL 1.50 0.00 196.5s
dart/reactive-cache ✔ PASS 1.50 1.50 226.7s
go/bank-account ✔ PASS 1.04 1.04 72.9s
go/dining-philosophers ✔ PASS 1.04 1.04 84.2s
go/errgroup-limit ✘ FAIL 1.14 0.00 122.7s
go/parallel-letter-frequency ✔ PASS 1.04 1.04 76.5s
go/react ✔ PASS 1.14 1.14 165.2s
go/singleflight ✔ PASS 1.28 1.28 118.0s
kotlin/channel-multiplexer ✘ FAIL 1.50 0.00 253.5s
kotlin/flow-processor ✔ PASS 1.50 1.50 239.4s
kotlin/lru-cache ✔ PASS 1.09 1.09 380.2s
rust/circular-buffer ✔ PASS 1.12 1.12 72.3s
rust/doubly-linked-list ✔ PASS 1.24 1.24 102.4s
rust/generational-arena ✘ FAIL 1.24 0.00 97.9s
rust/macros ✘ FAIL 1.50 0.00 91.6s
rust/parallel-letter-frequency ✔ PASS 1.12 1.12 46.4s
rust/regex-lite ✘ FAIL 1.40 0.00 61.7s
typescript/csv-lite ⚠ PARTIAL 1.36 1.02 122.1s
typescript/forth ✔ PASS 1.26 1.26 103.6s
typescript/glob ✔ PASS 1.14 1.14 66.9s
typescript/promise-pool ✘ FAIL 1.20 0.00 104.4s
typescript/react ✔ PASS 1.14 1.14 209.6s
zig/arena-allocator ✘ FAIL 1.50 0.00 246.3s
zig/comptime-json ✘ FAIL 1.50 0.00 245.0s
zig/small-vector ✔ PASS 1.34 1.34 116.5s

Verification

  • Harness Version: v1.6.1
  • Weight Version: 2.0
  • Tasks Hash: blake3:168696ded119da3ea973f15bde9c00151c127b3f54168196a84212827fa1b1f0
  • Results Hash: blake3:13a492d928fc92f7dee3a3fa89073e489556be6d6636084d76df35d0123e378c

Generated by SanityHarness on 2026-02-24T150330

Run Statistics

Tasks Passed 16 / 26
Duration 63m
Cost (Est) -
By Language
dart 33%
go 83%
kotlin 67%
rust 50%
typescript 80%
zig 33%