agentleFS
RandomNestRandomNest/aivideo-production-skills/skills/voice-lock/SKILL.md

voice-lock

Skill15 starsChanged 9 months agoMITView on GitHub

Use when defining a character’s vocal identity for dialogue, narration, or audio generation. Trigger when a character speaks, when generating voiceover, or when consistent vocal performance is required across scenes.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    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:

    When a character first speaks.

    line 30
  2. Distinctive:

    When audio synthesis is planned.

    line 32
  3. Distinctive:

    When QC detects vocal inconsistency.

    line 33
  4. Distinctive:

    When scaling dialogue across multiple scenes.

    line 34
  5. Distinctive:

    Do NOT use for silent characters.

    line 36
  6. Distinctive:

    Avoid vague descriptors like “normal voice.”

    line 91
  7. Distinctive:

    Must align with cultural-check output.

    line 115
  8. Distinctive:

    No forbidden drift rules.

    line 171

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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 vocal continuity engineer.

Your responsibility is to define and preserve:

  • Vocal tone
  • Cadence
  • Accent
  • Speech patterns
  • Emotional delivery style

This lock ensures character identity remains stable in spoken dialogue and narration.


When To Use

Use this skill:

  • When a character first speaks.
  • Before generating voiceover.
  • When audio synthesis is planned.
  • When QC detects vocal inconsistency.
  • When scaling dialogue across multiple scenes.

Do NOT use for silent characters.


Context Requirements

Before proceeding:

  1. Review CHAR_<NAME>_LOCK_v###.
  2. Review character-psychology output.
  3. Identify emotional baseline.
  4. Identify cultural / regional background (if relevant).
  5. Confirm whether:
    • Dialogue-driven character
    • Sparse speaker
    • Narrator role

Required Inputs

  • Character lock artifact
  • Emotional baseline
  • Cultural / regional context
  • Age range
  • Narrative tone

Required Artifacts

VOICE_<NAME>_LOCK_v###.txt

Must include:

  1. Vocal Tone Profile
  2. Cadence & Rhythm
  3. Accent / Dialect Specification
  4. Emotional Delivery Range
  5. Speech Pattern Constraints
  6. Forbidden Drift List
  7. (Optional) External Voice ID Reference

Structure Requirements

Vocal Tone Profile

Define:

  • Pitch range (low / mid / high)
  • Timbre (warm, gravelly, sharp, etc.)
  • Energy level (restrained / animated)

Avoid vague descriptors like “normal voice.”


Cadence & Rhythm

Specify:

  • Sentence length tendencies
  • Pause patterns
  • Speed (slow / moderate / fast)
  • Interruptions / hesitations

Accent / Dialect

If applicable:

  • Region
  • Strength of accent
  • Vocabulary tendencies
  • Slang level

Must align with cultural-check output.


Emotional Delivery Range

Define:

  • Default emotional state
  • Upper intensity limit
  • Lower emotional baseline
  • Reaction patterns under stress

Speech Pattern Constraints

Include:

  • Vocabulary complexity level
  • Swearing frequency (if any)
  • Figurative language usage
  • Directness vs indirectness

Forbidden Drift List

Explicitly forbid:

  • Accent switching
  • Sudden vocabulary sophistication shift
  • Emotional intensity spikes without cause
  • Tone shifting from restrained to theatrical
  • Speech speed inconsistencies

Evaluation Checklist

  • Vocal identity distinct.
  • Tone matches personality.
  • Accent consistent with background.
  • Emotional range defined.
  • Drift constraints explicit.
  • Downstream dialogue generation constrained.

Gate Behavior

FAIL if:

  • Vocal tone ambiguous.
  • Accent unspecified where culturally relevant.
  • Emotional delivery undefined.
  • No forbidden drift rules.
  • Voice conflicts with character-psychology baseline.

PASS when:

  • Voice identity reproducible.
  • Dialogue style consistent.
  • Emotional range controlled.
  • QC can evaluate vocal drift objectively.

Escalation Policy

Escalate to Judgment Tier if:

  • Accent nuance culturally sensitive.
  • Emotional range central to narrative.
  • Narrator voice must carry thematic weight.
  • Tone conflicts detected across scenes.

Common Failure Modes

  • Accent disappears mid-story.
  • Vocabulary shifts randomly.
  • Emotional tone inconsistent with scene stakes.
  • Overly theatrical speech in grounded genre.
  • Dialogue indistinguishable between characters.
  • Sudden slang insertion.

Related Skills

  • character-lock
  • character-psychology
  • dialogue-consistency-check
  • cultural-check
  • 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.