メニュー
Rewrite a draft message, explanation, or note to strip AI voice, slop, and filler while preserving meaning. Use to polish functional prose before sending.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Query Claude Code session history via a DuckDB index over `~/.claude/projects/`. Use when asked about Claude Code activity ("how many tokens today?", "what did I work on this week?") or instead of reading, grepping, or jq-ing session transcripts. Not for codebase search, git log queries, or arbitrary databases.
Triage and batch-implement Claude-tagged Things todos as PRs for the claude config repo, discover improvement candidates from session history, or watch open PRs to implement review feedback and close shipped todos. Use when the user wants to work on their Claude Code improvement backlog, process Things todos tagged claude-code, batch-implement configuration changes, mine session history for grounded config-change candidates (Discover mode), or watch this skill's open PRs for review feedback and merges (Watch mode).
Creating and optimizing Claude Code Skills including activation patterns, content structure, and development workflows. Use when creating new skills, converting memory files to skills, debugging skill activation, or understanding skill architecture and best practices.
Targeted interview for execution-time clarification. Use when you hit an ambiguity or decision point during implementation that needs user input before proceeding.
Comprehensive interview for planning new features or changes. Use when the user wants thorough upfront planning before implementation, or when starting work on a complex feature.
Curate the writing plugin's trope ruleset by auditing wordlist entries against session history and surfacing candidate phrases. Use when refreshing trope detection, reviewing wordlist health, or mining sessions for new AI-writing patterns to add or stale rules to remove.
| name | writing:rewrite |
| description | Rewrite a draft message, explanation, or note to strip AI voice, slop, and filler while preserving meaning. Use to polish functional prose before sending. |
| argument-hint | [file path or text; omit to read input from the clipboard] |
| user-invocable | true |
| context | fork |
| background | false |
| allowed-tools | ["Bash","Read"] |
Rewrite input text to match the style rules in references/style-rules.md. Preserve all functional information. Change only voice, word choice, and sentence structure.
$ARGUMENTS
Read input from one of these sources:
If $ARGUMENTS is a path to an existing file, read the file contents.
If $ARGUMENTS contains text, use it directly.
If $ARGUMENTS is empty, ask for the text to rewrite. You may offer to read it from the clipboard.
Before and after rewriting, run the scan script in single-input mode to find violations:
echo "<text>" | bun ${CLAUDE_SKILL_DIR}/../scan/scripts/scan.ts --input
Or pass a file path directly:
bun ${CLAUDE_SKILL_DIR}/../scan/scripts/scan.ts --input path/to/file.md
The script outputs one finding per line (line:col: category: message).
Fix every hard tell (em dash, copula avoidance, hedging, filler, vocabulary, and the other fixed-phrase categories), then re-run until those are clean.
Treat marketing verb findings as advisory. The script flags each one, but the live hook only objects when their weighted sum is high enough, so a lone marketing verb may be fine. Consider each in context and replace the ones that read as promotional rather than driving the count to zero.
Clear the lint findings as described above, then apply the structural rules from references/style-rules.md:
Do not add, remove, or restructure the content's meaning. Convey the same information in fewer, clearer words.
Display the rewritten text directly. Do not wrap it in a code block unless the input was code.
If the user asks, copy the result to the clipboard.