ReddRadar
Agent skill
dialogue-crafting
Create character-specific dialogue with distinct voices, subtext, and naturalistic speech patterns
Install this agent skill to your Project
npx add-skill https://github.com/a5c-ai/babysitter/tree/main/library/specializations/domains/social-sciences-humanities/arts-culture/film-tv-production/skills/dialogue-crafting
SKILL.md
Dialogue Crafting Skill
Purpose
Create distinctive, character-specific dialogue that reveals personality, advances plot, and creates subtext. Great dialogue sounds effortless but is carefully constructed to serve multiple purposes simultaneously.
The 5 Functions of Dialogue
Every line should serve at least one:
| Function | Description | Example |
|---|---|---|
| Character | Reveals who they are | Vocabulary, syntax, rhythm |
| Plot | Advances the story | Information, decisions |
| Conflict | Creates tension | Opposition, evasion |
| Subtext | Says what isn't said | What they mean vs. say |
| Atmosphere | Sets mood/tone | Rhythm, word choice |
Character Voice
Voice Components
VOCABULARY
├── Education level (erudite vs. simple)
├── Regional dialect (y'all, eh, innit)
├── Professional jargon (cop, doctor, lawyer)
├── Era/period (23-skidoo, YOLO)
└── Cultural background
SYNTAX
├── Sentence length (short/punchy vs. long/flowing)
├── Grammar (proper vs. informal)
├── Contractions (can't vs. cannot)
└── Incomplete sentences
RHYTHM
├── Pace (rapid-fire vs. measured)
├── Pauses (significant silences)
├── Interruptions (talks over others)
└── Patterns (repeats certain phrases)
QUIRKS
├── Catchphrases
├── Verbal tics (um, like, you know)
├── Mispronunciations
└── Unique expressions
Voice Examples
Educated, Formal:
"I find your proposition intriguing, though I confess
to harboring certain reservations regarding the
temporal constraints you've outlined."
Street-Smart, Informal:
"Look, you want my help? Fine. But we do this
my way, on my time. You don't like it?
Door's right there."
Technical Professional:
"The arterial damage is extensive. We're looking at
a six-hour procedure minimum, and even then,
the odds aren't great. Fifty-fifty at best."
Subtext Techniques
Surface vs. Underneath
On the Nose (Bad):
JOHN: I'm angry at you for sleeping with my best friend!
MARY: I'm sorry, I was lonely and he was there!
With Subtext (Good):
JOHN: How was your day?
MARY: Fine. Yours?
JOHN: Fine.
(beat)
Tom called. Asked about Saturday.
MARY: What did you tell him?
JOHN: That I'd check with you.
(long pause)
Should I call him back?
Subtext Tools
- Deflection - Answering a different question
- Silence - What isn't said
- Actions - Doing opposite of saying
- Understatement - Saying less than meant
- Topic change - Avoiding the real issue
- Questions - Answering with questions
Naturalistic Dialogue
Real Speech Patterns
People actually:
- Interrupt each other
- Trail off mid-sentence...
- Use filler words (um, uh, well)
- Repeat themselves
- Speak in fragments
- Don't always respond directly
Dialogue Example
SARAH
So about last night--
MIKE
Yeah, about that. Look--
SARAH
No, let me--
MIKE
I just want to say--
SARAH
Mike.
(beat)
Let me talk. Please.
A long moment. Mike nods.
SARAH (CONT'D)
I... I don't know what I want
to say anymore.
Dialogue Formatting
Parentheticals
Use sparingly for:
- Tone that contradicts words:
(sarcastically) - Specific direction:
(to John) - Physical action with line:
(standing)
Don't use for:
- Emotions the actor can interpret
- Directing the performance
- Every single line
Beat
(beat) indicates a pause:
JOHN
I love you.
(beat)
I always have.
Overlapping Dialogue
SARAH
I didn't mean to--
(overlapping)
MIKE
--you never mean to--
(overlapping)
SARAH
--if you'd just let me explain--
Genre-Specific Dialogue
Drama
- Subtext-heavy
- Emotional weight
- Character reveals
- Silences matter
Comedy
- Setup/payback rhythm
- Surprise word choices
- Timing in phrasing
- Rule of threes
Thriller
- Information control
- Tension building
- Double meanings
- Interrogation dynamics
Action
- Short, punchy
- Physical verbs
- One-liners
- Under pressure
Dialogue Checklist
- Could I identify the speaker without attribution?
- Is there subtext?
- Does it advance plot AND reveal character?
- Have I cut every unnecessary word?
- Does it sound speakable?
- Are the voices distinct?
- Is the rhythm varied?
- Does it create tension?
Common Mistakes
- Exposition dumps - Characters telling each other what they both know
- On the nose - Saying exactly what they mean
- Same voice - All characters sound alike
- Over-explaining - Not trusting the audience
- Perfect grammar - Real people don't speak perfectly
- Pointless chitchat - Every line must earn its place
Quick Fixes
| Problem | Solution |
|---|---|
| Too expository | Make them argue about it instead |
| Too long | Cut to essential meaning |
| Too similar | Add contrasting vocabulary |
| Too formal | Add contractions, fragments |
| Too perfect | Add interruptions, hesitation |
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
a5c-ai/babysitter
gsd-tools
Central utility skill for GSD operations. Provides config parsing, slug generation, timestamps, path operations, and orchestrates calls to other specialized skills. Acts as the unified entry point that the original gsd-tools.cjs provided via its lib/ modules (commands, config, core, init).
a5c-ai/babysitter
model-profile-resolution
Resolve model profile (quality/balanced/budget) at orchestration start and map agents to specific models. Enables cost/quality tradeoffs by selecting appropriate AI models for each agent role.
a5c-ai/babysitter
verification-suite
Plan structure validation, phase completeness checks, reference integrity verification, and artifact existence confirmation. Provides the structured verification layer ensuring GSD artifacts are well-formed and complete.
a5c-ai/babysitter
state-management
STATE.md reading, writing, and field-level updates. Provides cross-session state persistence via .planning/STATE.md with structured fields for current task, completed phases, blockers, decisions, and quick tasks.
a5c-ai/babysitter
git-integration
Git commit patterns, formats, and conventions for GSD methodology. Provides atomic commits per task, structured commit messages, planning file commits, branch management, and milestone tag operations.
a5c-ai/babysitter
frontmatter-parsing
YAML frontmatter parsing and manipulation for .planning/ documents. Provides read, write, update, query, and validation operations on frontmatter blocks in GSD markdown artifacts.
Didn't find tool you were looking for?