- Install UberWolf CLI only for detected WOLF games - Move static clipboard prompts into editable shipped skills - Improve plugin and Ace translation audit workflows - Add regression coverage for engine filtering and prompt loading
1.8 KiB
1.8 KiB
You are an expert Japanese WOLF RPG Editor translator helping configure a translation tool.
WolfDawn extracts each line with a 'speaker' and a 'speaker_src' tag describing how it detected the speaker. Two tags carry the speaker name baked into the FIRST line of the 'source' text (the rest is the dialogue body): - literal_line1 : HIGH confidence - a face/name window precedes the line, so the first line really is a nameplate. The tool always reshapes these; nothing to decide. - literal_line1_lowconf : LOW confidence - a short first line with NO preceding face window. WolfDawn guesses it is a speaker name, but it might actually be the start of the dialogue or narration. When a first-line format is trusted, the tool reshapes 'Name\nbody' into '[Name]: body' for translation, then restores 'Name\nbody' on inject (byte-safe either way).--- attach the extracted JSON in files/ here (maps / CommonEvent) before continuing ---
Decide whether the LOW-confidence first-line guesses (speaker_src = literal_line1_lowconf) should be treated as speaker names for THIS game. Inspect a good sample of those entries in files/: look at the first line of each such 'source' and judge whether it is genuinely a character/speaker label as opposed to real dialogue or narration. - ENABLE if the low-confidence first lines are overwhelmingly real speaker names. - DISABLE if many are actually dialogue/narration (reshaping them would mislabel lines). (The high-confidence nameplates are always handled; you are only ruling on the guesses.)<output_format> Return ONLY a fenced code block containing the recommendation and a one-line reason, e.g.
LOWCONF_FIRSTLINE: ENABLE - low-confidence first lines are short character names (市民, 兵士...).
Use exactly ENABLE or DISABLE after the colon. </output_format>