{
  "slug": "taste-skill",
  "title": "Taste Skill",
  "description": "A controlled three-pair evaluation of the pinned Taste Skill on a technical B2B landing-page redesign.",
  "status": "ADOPTED",
  "status_label": "Passed evaluation \u00b7 adopted",
  "package_hash": "aa194351b246b8b4799099d4ed7b033d29eab6e6e3d58d8d2172978be7b3ec89",
  "repo_url": "https://github.com/Leonxlnx/taste-skill/tree/e79ca9ec7e071eb3a3b623c4fb752e853fc3ed58/skills/taste-skill",
  "get_label": "Get the pinned skill",
  "model": "OpenAI GPT-5.6 Sol",
  "method": "Three blind baseline-versus-skill pairs in isolated Harbor containers. Same model, prompt, source page, and objective checker; only the skill installation changed.",
  "claim": "On this landing-page redesign, the skill won all three blind pairs and passed all three objective checks; no baseline did.",
  "status_explanation": "The skill passed Edge's preregistered three-pair adoption gate: a strict majority of blind wins, treatment verification passed, and treatment tool errors did not exceed baseline.",
  "stats": [
    {
      "value": "3 / 0 / 0",
      "label": "wins / ties / losses"
    },
    {
      "value": "3 of 3",
      "label": "skill outputs passed"
    },
    {
      "value": "0 of 3",
      "label": "baseline outputs passed"
    },
    {
      "value": "6",
      "label": "recorded agent runs"
    }
  ],
  "limitations": [
    "One task family with three paired samples; no confidence interval can support a catalog-wide efficacy claim.",
    "The skill explicitly excludes dashboards; this evaluation stays within its landing-page scope.",
    "No production conversion or user study was measured."
  ]
}