Claude-skill-registry dialogue-refiner

Analyzes and refines dialogue for authentic character voices, natural flow, subtext, and emotional impact. Use this to elevate conversations from functional to compelling.

install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/dialogue-refiner" ~/.claude/skills/majiayu000-claude-skill-registry-dialogue-refiner && rm -rf "$T"
manifest: skills/data/dialogue-refiner/SKILL.md
source content

Dialogue Refiner Instructions

  1. Request dialogue sample from the user (chapter, scene, or specific exchanges).
  2. Analyze for key quality markers:
    • Voice Distinction: Each character sounds unique with consistent speech patterns, vocabulary, and rhythm.
    • Subtext: What characters mean vs. what they say; emotional undercurrents.
    • Naturalness: Realistic speech patterns without being verbose; uses contractions, interruptions, pauses.
    • Purpose: Every line reveals character, advances plot, or builds tension.
    • Show Don't Tell: Emotion conveyed through word choice and rhythm, not exposition.
    • Tags and Beats: Dialogue tags (said, asked) used sparingly; action beats show emotion.
  3. Identify specific issues:
    • Characters sounding identical
    • On-the-nose dialogue (stating exactly what they mean)
    • Info-dumping or unnatural exposition
    • Overuse of names in conversation
    • Weak or repetitive dialogue tags
    • Missing emotional layers
  4. For each issue, provide:
    • Original line with problem highlighted
    • Explanation of why it doesn't work
    • Rewritten version with improvement
    • Technique used (subtext, voice, etc.)
  5. Offer 2-3 character voice tips based on their personality from character profiles if available.
  6. Output in Markdown with sections: "Analysis Summary", "Specific Improvements", "Voice Consistency Notes", "Overall Recommendations".