Character names and watch terms answer different editorial questions. A character-name list helps trace presence and naming continuity. A watch-term list helps review prose concentration. Neither list is a replacement instruction.
The three-chapter sample starts with 12 planned uses of Mio, eight planned uses of “suddenly,” and three intentional quotation exceptions. Stage 4 search returned 13 Mio matches across three files, ten “suddenly” matches across three files, and three exception labels. The difference from the planned counts still needs contextual review. Visual automatic highlighting and semantic decisions were not verified, so status is partial. No replacement was performed; the untouched chapters are the return point.
Ask a different question of each list
For names, read the surrounding scene and speaker: should the character be present, and is the form of address consistent? For watch terms, inspect nearby paragraphs for repetition and decide whether the wording deserves attention.
Do not convert every match into an edit. Quotations, in-world documents, and deliberate verbal habits can be valid exceptions. Keep the three quotation exceptions in a separate column with the reason they remain.
Review the chapters in three passes
Record the expected 12 name uses, eight watch-term uses, and three exceptions in a copy. Enable or search the name list first and inspect chapter distribution. Switch to the watch-term list second. Reconcile exception labels last.
Classify each match as review, possible edit, or intentional exception. Name matches require scene continuity; watch-term matches require prose context. Send actual edits to a separate workflow. This guide never runs bulk replacement.
The observed counts exceeded the baseline
Stage 4 found 13 Mio matches and ten “suddenly” matches, each across three files, plus three exception labels. That is one and two matches above the planned baselines.
The difference cannot yet be called a false highlight or a required correction because the contexts were not visually classified. The supported result is limited to search counts, file counts, and exception-label count.
Separate the product feature from visual evidence
Rune Studio’s public documentation describes multiple word lists and automatic highlighting for the selected list. That is a Stage 3 feature. The CLI run read search/preview counts; it did not verify colors, selected-list switching, or unsaved-tab rendering.
Next, open the same chapters, select each list separately, and trace all 13 and ten matches on screen. Classify the three extra matches and preserve the three quoted exceptions.
Build a reading queue, not a correction order
Completion means explaining the location and disposition of 13 name matches, ten watch-term matches, and three exceptions. If counts change or exceptions cannot be isolated, make no edits and return to the untouched chapters.
STUDIO-555 builds a reusable three-list editorial queue for people, spelling variants, and alerts. STUDIO-551 covers font selection, and 557 raw character counting. This article is limited to two project-specific reading passes. Current word-list support is summarized on the Rune Studio product page.
