Conciseness

Lean and efficient throughout: dense tables for check codes and fix actions, copy-paste command blocks, and only non-obvious domain knowledge (e.g. 'trailing whitespace in <source> is significant and editors strip it', 'Slovenian shares many words with Croatian and Serbian'). No padding or explanation of concepts Claude already knows; the brief intro's failure examples ('a dropped </b> that turns a whole dialog bold') complement rather than repeat the table's technical conditions.

Actionability

Fully executable guidance: exact script invocations with flags and exit statuses ('python3 "$A" --leak mogoče,prepričani,odložišče,počiščeno --skip sl,hr,sr,sr_Cyrl'), a per-finding fix table mapping each code to a concrete action, a git command to audit only PR-touched files, compile checks (xmllint, lrelease6 with temp .qm), and the full contribution-cleanup commit sequence. Not a 4 because there are no gaps — common cases are covered with copy-paste-ready commands.

Workflow Clarity

The audit → fix → re-audit loop is explicitly sequenced with a hard validation gate ('Repeat steps 2 and 3 until it prints total: clean (exit status 0). Do not stop on "mostly clean"') plus a separate compile-check step with expected outputs. This is a batch operation and the feedback loop is present and explicit, so the batch-operation cap does not apply; error recovery paths (revert to empty unfinished) are given per finding.

Progressive Disclosure

Bundled files are listed up front with one-level-deep links that exist on disk, each with a clear when-to-read signal ('read when a string passes the script but looks wrong, or when deciding whether a machine translation is salvageable'). The script's detailed check semantics are correctly pushed to the script itself, general translation workflow is delegated to the sibling skill, and the body keeps only the operational overview. Not a 4 because navigation and content split have no gaps.

Specificity

Lists multiple specific concrete actions — 'Audits QtPass localization/*.ts files for broken %1/%n placeholders, unbalanced HTML tags, lost & mnemonics, mixed-script and cross-language artifacts, then fixes and re-audits until clean' — with comprehensive coverage of what the skill does. No vague language; matches the anchor for multiple specific concrete actions.

Completeness

Explicitly answers both questions: the 'what' is the audit-fix-re-audit capability with named checks, and the 'when' is an explicit 'Use when...' clause with three concrete trigger scenarios. Clear 5; a 4 would require the 'when' to be less explicit or specific, which is not the case.

Trigger Term Quality

Natural trigger phrases users would actually say: 'reviewing translations', 'a Weblate PR', 'checking a machine-translated batch', 'chasing a placeholder mismatch or broken HTML in a .ts file'. Includes synonyms (translations, machine-translated batch) and the file extension (.ts), matching the comprehensive-coverage anchor rather than the 'a few natural terms missing' level below.

Distinctiveness Conflict Risk

The QtPass .ts-audit niche is distinct with its own triggers, but 'Use when reviewing translations or a Weblate PR' is broad enough to also match a general localization workflow skill (the body itself references a sibling qtpass-localization skill covering general translation review), creating minor overlap risk with a closely related skill. Not a 5 since the first trigger clause is not tied to .ts files.