Conciseness

The body is dense with non-obvious, tool-specific guidance (flag combinations, token/similarity thresholds, triage heuristics) and wastes almost nothing on concepts Claude already knows. Not 5 because the placeholder Before/After comment blocks ("// Before: same block in two places") and a few very long sentences (workflow step 5) could be trimmed.

Actionability

Copy-paste-ready commands ("npx jscpd --reporters ai --ignore-identifiers --min-tokens 70 <path>") and concrete named examples (computeCartTotal/computeBasketTotal, buildInvoice/buildCreditNote) with precise thresholds (0.8, 0.9, 10 lines) make the guidance highly executable. Not 5 because the strategy code blocks are illustrative comment pseudocode rather than executable examples, and that flexibility is not explicitly justified.

Workflow Clarity

Nine explicitly sequenced steps include a concrete validation step ("Re-run jscpd with the same flags to confirm the clone is eliminated and the dup% ... went down", including the --ignore-identifiers re-check for renamed clones), an "Always ensure" checklist (all call sites updated, tests pass), and reporting of skipped candidates with reasons. Not 4 because validation checkpoints and the feedback loop are explicit, not implicit.

Progressive Disclosure

Well-organized sections (Prerequisites, Workflow, Refactoring Strategies, Tips) with the full jscpd option reference deferred one level deep to the clearly signaled sibling skill ("See the [jscpd](../jscpd/SKILL.md) skill for full option reference"); no bundle files exist, and nothing in the body clearly belongs in one. Not 4 because the split between this skill's workflow and the jscpd skill's option reference is clean and easy to navigate.

Specificity

"Refactor exact, renamed and near-miss clones using extract function, parameterize, module, constant, or base class strategies" lists multiple concrete actions with comprehensive coverage of both the clone kinds and the refactoring strategies. Not 4 because the action list covers the domain comprehensively rather than having minor gaps.

Completeness

The 'what' is clear (a guided workflow with named strategies), but 'when' is only weakly implied by "detected by jscpd" — there is no explicit "Use when..." clause, which caps completeness at 3 per the judging guidelines. Not 4 because the trigger condition must be inferred rather than read.

Trigger Term Quality

"copy-paste duplication", "duplication" and "refactor" are phrases users naturally say, alongside the tool name jscpd. Not 5 because common variations like "duplicate code", "repeated code" or "DRY" are missing; not 3 because the natural phrases present go beyond a single generic keyword.

Distinctiveness Conflict Risk

"copy-paste duplication detected by jscpd" carves out a clear niche distinct from general code-quality skills. Not 5 because a request like "refactor this duplicated code" could also plausibly route to a general refactoring skill, giving minor overlap risk.