Add operational-value grading for Duplicate Code Detector

[Copilot]

Adds a deterministic operational-value grader for Duplicate Code Detector, measuring remediation of its duplicate-code reports without conflating workflow activity with repository outcomes.

• Frozen evidence contract

  • Binds reports to detector runs by generated run ID, with analyzed-commit fallback for historical reconstruction.
  • Uses only qualifying [duplicate-code] Duplicate Code: issues and their closing PR merge state.
  • Defines a 14-day evidence horizon, attainment-only baseline, and null handling for unavailable or absent evidence.

• Primary metric

  • Reports the fraction of qualifying duplicate-code reports resolved through a merged closing PR by maturity.

{
  "value": 0.5,
  "opportunityKey": "run:32073363134"
}

• Workflow integration
  • Registers the evaluator under graders.operational-value.
  • Regenerates the workflow lock file from the Markdown source.

Original prompt

[operational-value] duplicate-code-detector

Use the operational-value-designer skill to add and verify one deterministic
operational-value grader for github/gh-aw workflow duplicate-code-detector.

Requirements:

• Recover adoption-time intent and pre-adoption evidence independently for this workflow. Do not copy Daily File Diet's metric or evidence contract.
• If no direct operational metric can be defended from accepted repository evidence, stop and explain the blocker. Do not substitute execution quality, output volume, traces, or agent judgment.
• Change only .github/workflows/duplicate-code-detector.md, its generated lock file, .github/graders/duplicate-code-detector-operational-value.sh, and reports/operational-value/duplicate-code-detector-operational-value.{json,md,svg} unless a directly necessary test or documentation fix is required.
• Preserve one absolute primary value in [0,1]. Keep diagnostics separate and treat missing evidence as null, never zero.
• Support historical case reconstruction so pre-grader runs can be reported.
• Do not edit source-managed workflow files or generated lock files by hand.

Validation and report:

1. Run .github/skills/operational-value-designer/scripts/verify-operational-value-evaluator.sh .github/graders/duplicate-code-detector-operational-value.sh
2. Run gh aw compile .github/workflows/duplicate-code-detector.md
3. Run gh aw graders operational-value report duplicate-code-detector --repo github/gh-aw
4. Review the generated JSON, Markdown, and SVG for coverage failures or misleading aggregation before finishing.

Keep this pull request limited to one workflow so it can be reviewed and merged independently.

[github-actions]

Category: feature | Risk: medium | Score: 50 (impact 28 + urgency 12 + quality 10) | Action: batch_review | Batch: operational-value-grading

Generated by 🔧 PR Triage Agent · copilot · mai10 · 22.2 AIC · ⌖ 2.85 AIC · ⊞ 15.1K · ◷

[github-actions]

Category: feature | Risk: medium | Score: 50 (impact 28 + urgency 12 + quality 10) | Action: batch_review | Batch: operational-value-grading

Generated by 🔧 PR Triage Agent · copilot · mai10 · 22.2 AIC · ⌖ 2.85 AIC · ⊞ 15.1K · ◷