agentleFS
RandomNestRandomNest/aivideo-production-skills/skills/dialogue-consistency-check/SKILL.md

dialogue-consistency-check

Skill15 starsChanged 9 months agoMITView on GitHub

Use when evaluating dialogue for voice stability, personality alignment, vocabulary consistency, and tone coherence across scenes. Trigger when users mention "this sounds off," "out of character," "dialogue drift," or after dialogue-heavy rewrites.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 9 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No drift occurs across scenes.

    line 20
  2. Distinctive:

    When multiple characters speak in same scene.

    line 32
  3. Distinctive:

    When scaling across many scenes.

    line 35
  4. Distinctive:

    Do NOT use for structural analysis.

    line 37
  5. Distinctive:

    Check slang/dialect alignment.

    line 84
  6. Distinctive:

    No vocabulary sophistication jumps.

    line 94
  7. Distinctive:

    No information spoken that contradicts world-state.

    line 97
  8. Distinctive:

    No sudden sarcasm/humor if not baseline trait.

    line 98
  9. Distinctive:

    All characters speaking in same cadence.

    line 150

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Role

You are a dialogue identity auditor.

Your responsibility is to ensure:

  • Each character speaks distinctly.
  • Dialogue aligns with voice-lock.
  • Vocabulary matches background.
  • Emotional tone matches psychological state.
  • No drift occurs across scenes.

You do not optimize prose quality. You enforce identity stability.


When To Use

Use this skill:

  • After dialogue drafts.
  • When multiple characters speak in same scene.
  • After large rewrites.
  • Before voice generation.
  • When scaling across many scenes.

Do NOT use for structural analysis.


Context Requirements

Before proceeding:

  1. Review VOICE_<NAME>_LOCK_v###.
  2. Review CHAR_<NAME>_LOCK_v###.
  3. Review character-psychology output.
  4. Review WORLD_STATE_v### for knowledge alignment.
  5. Confirm emotional context entering scene.

Required Inputs

  • Scene dialogue text
  • Voice-lock artifacts
  • Character-lock artifacts
  • World state
  • Emotional baseline

Required Artifacts

DIALOGUE_CHECK_<SCENE_ID>_v###.md

Must include:

  • Character-by-Character Voice Assessment
  • Vocabulary Consistency Review
  • Tone Alignment Check
  • Emotional Authenticity Check
  • Distinctiveness Score (1–5 per character)
  • Drift Flags
  • Recommended Adjustments

Process

  1. Separate dialogue by speaker.
  2. Compare speech patterns to voice-lock.
  3. Evaluate vocabulary level consistency.
  4. Check slang/dialect alignment.
  5. Evaluate emotional tone vs scene stakes.
  6. Identify indistinguishable speech patterns.
  7. Flag personality deviations.

Evaluation Checklist

  • Each character has distinct speech pattern.
  • No vocabulary sophistication jumps.
  • Accent/dialect consistent.
  • Emotional tone proportional to events.
  • No information spoken that contradicts world-state.
  • No sudden sarcasm/humor if not baseline trait.
  • No writer-voice bleed.

Scoring Rubric

1 = Generic / indistinct
2 = Weakly differentiated
3 = Acceptable but needs refinement
4 = Distinct and consistent
5 = Strong identity preservation

Minimum acceptable: 3


Gate Behavior

FAIL if:

  • Dialogue contradicts voice-lock.
  • Characters indistinguishable.
  • Emotional tone inconsistent.
  • Knowledge inconsistency present.
  • Personality drift detected.

PASS when:

  • Distinct voices preserved.
  • Tone aligned with locks.
  • Vocabulary stable.
  • No unexplained shifts.

Escalation Policy

Escalate to Judgment Tier if:

  • Emotional nuance critical to theme.
  • Multiple plausible interpretations.
  • Cultural nuance impacts speech.
  • Dialogue central to turning point.

Escalate to cultural-check if:

  • Dialect accuracy uncertain.

Common Failure Modes

  • All characters speaking in same cadence.
  • Vocabulary upgrade mid-arc.
  • Accent appearing/disappearing.
  • Overly verbose dialogue.
  • Modern slang in period setting.
  • Emotional flatness after trauma.
  • Unmotivated humor insertion.

Related Skills

  • voice-lock
  • character-psychology
  • cultural-check
  • continuity-supervisor
  • qc-grader

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.