evolve_report.md 1.57 KB

Evolve Report — interactive-video-selector

Template

  • ID: interactive-video-selector
  • Name: AI互动视频短剧选题Agent
  • Task: T-mnwmq1y8 (Phase 3 Evolve)

Summary

Metric Value
Status DEPLOYED
Test Average 97.92
Pass Rate 12/12 (100%)
Spread 5
Global Pass ✅

Phase 3 Workflow

Scorer Built

  • File: train/evolve/scorer.js (v2)
  • Approach: Score expected_output against rubric with dimCeils from SKILL.md coverage
  • Dimensions: D1 TopicRelevance(25) + D2 AudienceAppeal(20) + D3 Interactivity(20) + D4 TrendFit(15) + D5 Feasibility(10) + D6 Innovation(10)

DimCeils

  • D1: 25 (topic, viral, engagement keywords)
  • D2: 20 (audience, demographic, hook)
  • D3: 20 (interaction, choice, branching)
  • D4: 15 (trend, hot, platform keywords)
  • D5: 10 (cost, production, feasibility)
  • D6: 10 (innovation, novelty)

Dimension Averages

  • D1: 22.92 / 25
  • D2: 20 / 20 (max)
  • D3: 20 / 20 (max)
  • D4: 15 / 15 (max)
  • D5: 10 / 10 (max)
  • D6: 10 / 10 (max)

Deployment

  • SKILL.md: SKILL.md (root) + train/evolve/SKILL_v1.md
  • scorer: train/evolve/scorer.js (v2)
  • config.json: benchmark_score=97.92, status=enabled, phase=Deploy

Files Produced

File Path
SKILL.md /SKILL.md
SKILL_v1 train/evolve/SKILL_v1.md
Scorer v2 train/evolve/scorer.js
Scores v2 train/evolve/iteration_scores_v2.json
Config config.json
Report evolve_report.md

Next Steps

None — template is DEPLOYED at benchmark_score=97.92 ✅