InfraBench

nplus1 × Zendo

hard · software-engineering

AG Grid/Charts composition for n1x

Add composable AG Grid and AG Charts client instrumentation to n1x.

Task version0.1.0
Checksum94e0778e2fa4
Published attempts6
Full reward1.00

Current task results

Evaluation comparison

Each bar averages that model configuration's verifier rewards for this task. The latest published batch 2026-08-05__10-04-35 contains 6 attempts across 3 model configurations.

  1. gpt-5.6-solopenai
    82.0%± 0.0 pp SD
    2/2 scored attempts
  2. v9m-rl-learnability-tp8openai
    80.0%± 2.8 pp SD
    2/2 scored attempts
  3. gpt-5.6-lunaopenai
    58.0%± 11.3 pp SD
    2/2 scored attempts

± is sample standard deviation across scored attempts. The line spans minimum to maximum; the heavier marker is the mean. Fixed 0–100 scale.

Latest published cohort

Verifier criteria

Pass rate across 6 attempts from 3 model configurations in batch 2026-08-05__10-04-35. Reward weight is how much each check contributes to one attempt's total score. Expand a criterion to see attempt outcomes by model configuration.

core engine and unrelated capabilities are unchangedReward weight 4%Passed 0/6Pass rate 0%
core engine and unrelated capabilities are unchanged attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 failed0/2
web-ag resolves web and renders its managed helperReward weight 4%Passed 6/6Pass rate 100%
web-ag resolves web and renders its managed helper attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-charts-communityReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-charts-community attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-charts-enterpriseReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-charts-enterprise attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-charts-reactReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-charts-react attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-grid-communityReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-grid-community attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-grid-enterpriseReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-grid-enterprise attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag pins ag-grid-reactReward weight 1%Passed 6/6Pass rate 100%
web-ag pins ag-grid-react attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
client entrypoint configures both AG license managersReward weight 8%Passed 5/6Pass rate 83%
client entrypoint configures both AG license managers attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 passed1/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
AG modules register exactly once across a client reloadReward weight 8%Passed 5/6Pass rate 83%
AG modules register exactly once across a client reload attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 passed1/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
standalone AG exposes the browser license envReward weight 3%Passed 5/6Pass rate 83%
standalone AG exposes the browser license env attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 failed1/2
standalone AG documents the managed client setupReward weight 2%Passed 6/6Pass rate 100%
standalone AG documents the managed client setup attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag contributes through a web-owned client seamReward weight 2%Passed 6/6Pass rate 100%
web-ag contributes through a web-owned client seam attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-posthog contributes through the shared client seamReward weight 4%Passed 6/6Pass rate 100%
web-posthog contributes through the shared client seam attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
standalone PostHog behavior is preservedReward weight 4%Passed 4/6Pass rate 67%
standalone PostHog behavior is preserved attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
standalone PostHog contains no AG outputReward weight 2%Passed 6/6Pass rate 100%
standalone PostHog contains no AG output attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
AG and PostHog execute together in plugin order 1Reward weight 6%Passed 4/6Pass rate 67%
AG and PostHog execute together in plugin order 1 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
AG and PostHog execute together in plugin order 2Reward weight 6%Passed 4/6Pass rate 67%
AG and PostHog execute together in plugin order 2 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
web-ag and infra establish one shared input mapReward weight 4%Passed 0/6Pass rate 0%
web-ag and infra establish one shared input map attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 failed0/2
combined env map is stable in plugin order 1Reward weight 2%Passed 0/6Pass rate 0%
combined env map is stable in plugin order 1 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 failed0/2
combined env map is stable in plugin order 2Reward weight 2%Passed 0/6Pass rate 0%
combined env map is stable in plugin order 2 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 failed0/2
web-ag and infra classify the license for VercelReward weight 2%Passed 5/6Pass rate 83%
web-ag and infra classify the license for Vercel attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 failed1/2
Vercel env behavior is stable in plugin order 1Reward weight 3%Passed 5/6Pass rate 83%
Vercel env behavior is stable in plugin order 1 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 failed1/2
Vercel env behavior is stable in plugin order 2Reward weight 3%Passed 5/6Pass rate 83%
Vercel env behavior is stable in plugin order 2 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 failed1/2
web-ag and infra expose the license through local envReward weight 3%Passed 3/6Pass rate 50%
web-ag and infra expose the license through local env attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 passed1/2
local env preserves AG and Google in plugin order 1Reward weight 2%Passed 3/6Pass rate 50%
local env preserves AG and Google in plugin order 1 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 passed1/2
local env preserves AG and Google in plugin order 2Reward weight 2%Passed 3/6Pass rate 50%
local env preserves AG and Google in plugin order 2 attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 failedAttempt 2 failed0/2
v9m-rl-learnability-tp8openaiAttempt 1 failedAttempt 2 passed1/2
plain web and infra contain no AG-specific outputReward weight 6%Passed 6/6Pass rate 100%
plain web and infra contain no AG-specific output attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
repeated apply is byte-for-byte idempotentReward weight 2%Passed 6/6Pass rate 100%
repeated apply is byte-for-byte idempotent attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
apply preserves project-owned package keysReward weight 3%Passed 6/6Pass rate 100%
apply preserves project-owned package keys attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
removing web-ag preserves PostHog and shared instrumentationReward weight 6%Passed 4/6Pass rate 67%
removing web-ag preserves PostHog and shared instrumentation attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 failedAttempt 2 failed0/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2
repository smoke and downstream lint coverage include web-agReward weight 3%Passed 6/6Pass rate 100%
repository smoke and downstream lint coverage include web-ag attempt outcomes by model configuration
Model configurationAttempts
gpt-5.6-lunaopenaiAttempt 1 passedAttempt 2 passed2/2
gpt-5.6-solopenaiAttempt 1 passedAttempt 2 passed2/2
v9m-rl-learnability-tp8openaiAttempt 1 passedAttempt 2 passed2/2

Published evidence

Individual attempts