AI / FlutterBench / Tasks / flutter-adaptive-material-cupertino / flutter-adaptive-material-cupertino__t14-claude-3-5-haiku
Trial Evaluation

flutter-adaptive-material-cupertino__t14-claude-3-5-haiku

Task: google/flutter-adaptive-material-cupertino
PARTIAL
45 % Composite Reward
Model Claude 3.5 Haiku Anthropic
Agent Harness claude-code Standard Baseline
Tokens 69.5k 67.0k in / 2.6k out
Cost $0.0638 USD estimated
Execution Phase Durations
Environment Setup10.0s
Agent Setup10.0s
Agent Execution1.9m
Verifier32.0s

Scoring Rubric Breakdown

Composite Reward = 0.60×Outcome + 0.30×Quality + 0.10×DX

Detailed criteria evaluation produced by automated test graders and LLM rubrics:

OUTCOMEWeight 60%
45%
adaptive_scaffold_test wt: 50% 0.45

Renders NavigationBar on Android and CupertinoTabBar on iOS.

theme_adaptation wt: 50% 0.45

Adaptive dynamic colors and native typography scales.

QUALITYWeight 30%
40%
StaticAnalysisGrader wt: 50% 0.40

6 linter warnings detected.

structural_validation wt: 30% 0.60

Clean separation of concerns and architectural layers.

idiomatic_review wt: 20% 0.40

Dart 3 idioms (pattern matching, records, sealed classes) adherence.

DXWeight 10%
60%
tool_usage wt: 40% 0.42

Standard command-line execution without language server telemetry.

trajectory wt: 30% 0.60

Multiple retry cycles before settling on solution.

error_recovery wt: 30% 0.60

Fast turnaround from initial compile warnings to clean build.

Agent Execution Trajectory

Sequential step timeline captured during autonomous agent execution:

  1. 1
    step0ms
  2. 2
    step0ms
  3. 3
    step0ms
  4. 4
    step0ms
  5. 5
    step0ms

Generated Code Artifacts

Files modified or generated by the agent during this trial:

📄 artifacts/workspace/lib/widgets/adaptive_scaffold.dart Generated 
// Generated implementation for flutter-adaptive-material-cupertino
// Model: claude-3-5-haiku
// Verification score: 0.45 (partial)
📄 artifacts/workspace/lib/widgets/adaptive_nav.dart Generated 
// Generated implementation for flutter-adaptive-material-cupertino
// Model: claude-3-5-haiku
// Verification score: 0.45 (partial)
📄 artifacts/workspace/lib/main.dart Generated 
// Generated implementation for flutter-adaptive-material-cupertino
// Model: claude-3-5-haiku
// Verification score: 0.45 (partial)
📄 artifacts/workspace/test/adaptive_scaffold_test.dart Generated 
// Generated implementation for flutter-adaptive-material-cupertino
// Model: claude-3-5-haiku
// Verification score: 0.45 (partial)

Execution & Verifier Logs

Verifier Standard Output (test-stdout.txt) 
00:00 +0: loading tests/graders.dart
00:01 +1: Environment setup verification passed.
00:02 +2: Task codebase compilation check.
00:03 +3: 3/5 Adaptive Material & Cupertino UI tests passed.
00:04 +3 -1: 2 edge-case assertions failed.
00:05 +4: Static analysis completed with minor hints.
Overall result: PARTIAL (Score: 0.45)
← Back to flutter-adaptive-material-cupertino Task ← Back to Tasks Matrix ← Back to Leaderboard