{
  "title": "Resumize resume-generation internal evaluation",
  "recordedAt": "2026-07-22T20:42:46.722Z",
  "source": "Saved local baseline used by the workflow eval harness; not a new CI run",
  "dataset": "resume-gen-v1",
  "cases": 37,
  "completed": 37,
  "pipelineErrors": 0,
  "checksPassed": 443,
  "checksTotal": 444,
  "recordedModelMap": {
    "resumize": "gpt-5-mini",
    "resumizePlus": "gpt-5.4-mini",
    "resumize-plus": "gpt-5.4-mini",
    "skillMatcher": "gpt-5-mini"
  },
  "sourceRunCommit": "af9a10681ba1633143b6bf17a8650608cb2d6e29",
  "sourceFileSha256": "6d074ae30d4f1c17eecab22e5337941b441be5275d3c0048d2831061dc455573",
  "checks": [
    {
      "id": "schema.jobAnalysis",
      "label": "Job-analysis structure",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "schema.experienceMapping",
      "label": "Experience-mapping structure",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "schema.resumeContent",
      "label": "Resume-content structure",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "merge.mapping",
      "label": "Experience identity in mapping",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "merge.integrity",
      "label": "Experience identity in final content",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "grounding.skills",
      "label": "Skill source-match check",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "grounding.dates",
      "label": "Date merge-match check",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "bullets.count",
      "label": "Bullet-count limits",
      "pass": 36,
      "fail": 1
    },
    {
      "id": "keywords.coverage",
      "label": "Keyword-coverage check",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "style.noDashes",
      "label": "Dash-style check",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "style.summaryLength",
      "label": "Summary-length check",
      "pass": 37,
      "fail": 0
    },
    {
      "id": "style.noPlaceholders",
      "label": "Placeholder-pattern check",
      "pass": 37,
      "fail": 0
    }
  ],
  "limits": [
    "Internal, historical evaluation of English-language content generation; not independently audited or a current-version certification.",
    "37 anonymized profile/job cases; some profiles contribute two cases. Dataset spans seven industry groups and three seniority groups, with eight sparse profiles.",
    "Certifications and projects were not available in these dataset snapshots.",
    "Checks measure specified structure, matching and style rules, not overall resume quality. A passing skill source-match check does not prove every skill is supported by the candidate; job text and analysis also count as sources.",
    "Keyword matching is not an ATS acceptance test. No AI judge results were recorded in this baseline.",
    "PDF layout, parsing in employer ATS systems, competitor performance and hiring outcomes were not tested.",
    "One case failed the bullet-count check. Later checks added to the harness are not included in this snapshot."
  ]
}