distill-stylelisted

# /distill-style — Turn raw style-feedback entries into rule edits The companion to `/fetch-annotations` Step 7. That step *harvests* coauthor feedback as raw entries; this skill *distills* accumulated raw entries into proposed edits to the writing-style rule files, with user approval. ## Arguments - `/distill-style` — process all `Status: raw` entries in `research/rules/writing_feedback_log.md`. - `/distill-style --since YYYY-MM-DD` — only entries appended after that date. - `/distill-style --tag <tag>` — only one tag's cluster. ## Targets The rules live in three places. Decide per cluster: 1. `research-kit/rules/writing_style.md` — general baseline (public, sibling repo). Edit when the rule is general-purpose empirical-econ writing advice that any project would benefit from. 2. `research/rules/writing_style.md` — workspace overlay (banned phrases, referee carve-out, workspace-specific tics). Edit when the rule is workspace- or author-specific. Banned-phrase swaps land here. 3. `research-kit/rules/writing_style/<section>.md` — per-section files (`intro.md`, `methods.md`, `results.md`, `data.md`, `discussion.md`, `conclusion.md`, `theory.md`, `institutions.md`, `abstract.md`, `body.md`, `appendix.md`). Edit only when the entries explicitly address a section-specific concern. **Default = general.** Route to a per-section file only when the coauthor's note explicitly speaks to a section-level concern (e.g., "the intro doesn't motivate the outcom