Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Scriptwriter

ASecurity

Write production-ready video ad scripts with full production notes — scene-by-scene breakdowns with dialogue, B-roll, on-screen text, camera direction, hooks, and timing. Outputs individual script files plus a combined approval overview. Use when user says "/scriptwriter", "write a video script", "video ad script", "reels script", "Meta ad script", "scene-by-scene", "écris un script vidéo", "script de pub vidéo", "script reels", "rédige un script", "storyboard de pub".

2 stars
0 votes
0 copies
0 views
Added 10/2/2026
ai-agentsrustgoreactrailsgitperformance

Works with

cli

Security Analysis

A100/100

Scanned 10/2/2026

$npx -y skills add agentik-os/claude-code-skills --skill scriptwriter --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Scriptwriter?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Scriptwriter
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/agentik-os-scriptwriter/badge)](https://www.skillsdirectory.com/skills/agentik-os-scriptwriter)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: scriptwriter
description: Write production-ready video ad scripts with full production notes — scene-by-scene breakdowns with dialogue, B-roll, on-screen text, camera direction, hooks, and timing. Outputs individual script files plus a combined approval overview. Use when user says "/scriptwriter", "write a video script", "video ad script", "reels script", "Meta ad script", "scene-by-scene", "écris un script vidéo", "script de pub vidéo", "script reels", "rédige un script", "storyboard de pub".
---

# Scriptwriter

Write production-ready video ad scripts for social/Meta campaigns.

> **Portability note:** Output paths below use `~/clawd/output/{project}-campaign/` (an OmegaOS VPS convention). On any other machine, write to the project's own output dir (e.g. `./output/{project}-campaign/` or wherever the brief specifies) — keep the same internal structure.

## Single-voice craft (do NOT parallelize)

Scriptwriting is **one coherent voice, one creative judgment**. Do NOT fan this out to parallel agents, do NOT run an "audit/verify 2-of-3" loop, and do NOT split scenes across workers — a script's hook, emotional arc, and CTA must be authored by a single mind so rhythm and tone stay continuous. Write all scripts in a campaign sequentially, in the same hand.

**Craft upgrades (apply every time):**
1. **One ear, read aloud.** Every line is performed, not read. Mentally voice each scene at 2.5–3 words/sec; cut any phrase that trips the tongue. Contractions, short clauses, one idea per breath.
2. **Hook owns the first 3 seconds, the arc owns the rest.** Pick ONE hook formula (below) and commit — don't hedge two openings. Then make every scene earn the next: each beat raises a question the following beat answers, so the viewer can't scroll.
3. **Specificity beats adjectives.** Concrete number, named tool, real before/after > "amazing", "powerful", "game-changing". Voice-match the brief's brand; if no brand voice is given, default to plain, confident, conversational — never ad-speak.

## Output Structure

### Individual Script Files (for production)
```
~/clawd/output/{project}-campaign/
├── scripts/
│   ├── 01-{script-name}.md
│   ├── 02-{script-name}.md
│   └── ...
```

### Combined Overview (for approval)
```
├── scripts-overview.md    # All scripts in one doc
```

## Script Template

Each script file follows this structure:

```markdown
# {Brand} Video Ad Script: {Title}

**Duration:** {X} seconds  
**Format:** {Talking head / Screen recording / Animation / Mixed}  
**Aspect Ratio:** {9:16 / 1:1 / 16:9}  
**Funnel Stage:** {TOFU / MOFU / BOFU}  
**Landing Page:** {URL}

---

## 📋 Overview

| Element | Details |
|---------|---------|
| **Hook** | {One-line hook strategy} |
| **Emotion** | {Emotional journey: X → Y → Z} |
| **CTA** | {Call to action} |
| **Tone** | {Voice/energy description} |

---

## 🎬 SCENE 1: HOOK (0:00 - 0:03)

### Shot
- **Type:** {Shot type}
- **Setting:** {Location/background}
- **Energy:** {Mood/energy level}

### Dialogue
> "{Exact words}"

### On-Screen Text
```
{Text that appears on screen}
```

### B-Roll Options
1. {Option 1}
2. {Option 2}

### Direction
- {Camera/performance direction}
- {Timing notes}

---

[Continue for each scene...]

---

## 📝 FULL SCRIPT (Clean Copy)

{Complete script without formatting for teleprompter}

**Word count:** {N} words  
**Reading time:** ~{N} seconds

---

## 🎨 THUMBNAIL CONCEPT

{2-3 thumbnail options with descriptions}

---

## 🎥 PRODUCTION NOTES

{Equipment, recording tips, editing notes}

---

## 📊 VARIATIONS TO TEST

| Element | A | B |
|---------|---|---|
| {Element} | {Version A} | {Version B} |

---

## ✅ PRE-PUBLISH CHECKLIST

- [ ] Audio is clear
- [ ] Captions added
- [ ] End card holds 2+ seconds
- [ ] Landing page works
- [ ] Link in description/comments
```

## Script Structure by Duration

### 15-Second Script (Stories/Reels)
```
[0-2s]  HOOK - Pattern interrupt
[2-8s]  VALUE - One key insight
[8-13s] CTA - Clear next step
[13-15s] END CARD
```

### 25-30 Second Script (Standard)
```
[0-3s]   HOOK - Pattern interrupt
[3-10s]  PROBLEM - Agitate pain point
[10-20s] SOLUTION - Present offer
[20-27s] PROOF + CTA - Social proof + action
[27-30s] END CARD
```

### 45-60 Second Script (Long-form)
```
[0-5s]   HOOK - Strong opening
[5-15s]  PROBLEM - Deep agitation
[15-30s] SOLUTION - Explain transformation
[30-45s] PROOF - Testimonials/results
[45-55s] CTA - Clear call to action
[55-60s] END CARD
```

## Hook Formulas

### Pattern Interrupt
> "If you're still [wrong behavior], you're [consequence]."

### Curiosity Gap
> "I [achieved result] and here's what nobody tells you..."

### Controversial Take
> "Stop [common advice]. Here's what actually works."

### Identity Call-Out
> "This is for [specific person] who [specific situation]..."

### Question Hook
> "Why do [surprising thing] when you could [better alternative]?"

### Result Lead
> "[Company] went from [before] to [after]. Here's how."

## On-Screen Text Guidelines

1. **Keep it short** — 3-5 words max per text element
2. **Reinforce, don't repeat** — Add to dialogue, don't duplicate
3. **Use hierarchy** — Important words bigger/bolder
4. **Time it right** — Appear with dialogue, not before
5. **Position consistently** — Same area throughout video

### Text Styles
```
EMPHASIS: All caps, bold
"Quote style": In quotation marks
Stat/number: Large, accent color
CTA: Arrow or emoji pointing (👇 ➡️)
```

## B-Roll Categories

### Screen Recordings
- Tool interfaces (ChatGPT, Make, etc.)
- Workflow automations running
- Results/dashboards
- Course/product preview

### Talking Head Variations
- Different angles (straight, 3/4)
- Different framings (tight, medium, wide)
- Gestures and reactions

### Stock/Generic
- Person working at laptop
- Team collaboration
- Frustration → success moments
- Abstract tech visuals

### Brand Specific
- Product screenshots
- Customer testimonials (clips)
- Behind-the-scenes
- Event footage

## Emotional Arcs

### TOFU (Awareness)
```
Frustration → Curiosity → Hope
"I'm stuck" → "Wait, what?" → "I could do this"
```

### MOFU (Consideration)
```
Skepticism → Understanding → Confidence
"Does this work?" → "This makes sense" → "I need this"
```

### BOFU (Decision)
```
Hesitation → Trust → Urgency
"Is it worth it?" → "Others got results" → "I should act now"
```

## Pacing Guidelines

| Duration | Words | Pace |
|----------|-------|------|
| 15s | 35-45 | Fast, punchy |
| 30s | 65-80 | Conversational |
| 45s | 100-120 | Room to breathe |
| 60s | 140-160 | Storytelling pace |

**Rule of thumb:** 2.5-3 words per second for natural delivery.

## Workflow

### 0. Optional: analyze reference videos with /watch

When the brief includes reference or competitor videos (URL or local file), run /watch on each one first. The report.md gives you scene-by-scene pacing (cuts per minute, shot lengths), the word-level transcript of the opening, and hook frames. Treat it as the script research: mirror the pacing that holds attention, note where the reference hooks in the first 3 seconds, and steal the structure, not the words. Skip when the brief has no reference videos.

### 1. Receive Brief
From `/campaign_planner` or direct request:
- Ad concept and funnel stage
- Target duration
- Key messaging points
- Landing page
- Brand voice guidelines

### 2. Write Script
- Start with hook (most important)
- Build emotional arc
- Time out each section
- Add B-roll suggestions
- Include on-screen text

### 3. Review & Polish
- Read aloud for timing
- Check word count vs duration
- Ensure CTA is clear
- Verify landing page alignment

### 4. Add Production Notes
- Thumbnail concepts
- A/B test variations
- Equipment suggestions
- Editing notes

### 5. Deliver
- Save individual script file
- Add to overview document
- Send for approval

## File Naming Convention

```
{NN}-{slug}.md

01-free-crash-course.md
02-ai-readiness-quiz.md
03-ai-operator-course.md
04-team-training-case-study.md
```

## Overview Document Structure

```markdown
# {Brand} Campaign Scripts Overview

**Campaign:** {Name}
**Total Scripts:** {N}
**Date:** {YYYY-MM-DD}

---

## Summary

| # | Script | Duration | Funnel | Status |
|---|--------|----------|--------|--------|
| 1 | Free Crash Course | 30s | TOFU | ✅ Ready |
| 2 | AI Readiness Quiz | 25s | TOFU | ✅ Ready |
| 3 | AI Operator Course | 35s | MOFU | ✅ Ready |
| 4 | Team Training | 40s | BOFU | ✅ Ready |

---

## Script 1: Free Crash Course

[Full script content]

---

## Script 2: AI Readiness Quiz

[Full script content]

---

[Continue for all scripts...]
```

## Integration

Works with:
- `/campaign_planner` — Provides creative briefs
- `/creative_director` — Orchestrates full asset creation
- `/ad_designer` — Creates companion image ads
- `/page_designer` — Creates landing pages scripts point to

## Output Contract

Every invocation MUST produce:
1. **One file per script** at `scripts/{NN}-{slug}.md`, each following the Script Template (overview table, scene-by-scene shots/dialogue/on-screen text/B-roll/direction, clean full-script copy, thumbnail, production notes, A/B variations, pre-publish checklist).
2. **One `scripts-overview.md`** with the summary table + every full script inlined, for one-pass approval.
3. A short **delivery note** listing the files written + their absolute paths.

## Verify (before claiming done)

Run this self-check on every script before reporting:
- **Timing:** word count ÷ ~2.75 ≈ stated duration (within ±15%). If off, trim or pad — never mislabel.
- **Structure:** scene timestamps sum to the total duration with no gap/overlap; END CARD present.
- **Files:** every script file exists AND is mirrored in `scripts-overview.md`; naming matches `{NN}-{slug}.md`.
- **CTA + landing page:** present, specific, and the URL matches the brief (do not invent one).
- Read the clean full-script copy aloud once; if any line snags, rewrite it.

## Guardrails (evidence, no hallucination, scope)

- **No invented facts.** Stats, testimonials, results, prices, feature claims, and the landing-page URL come from the brief. If a needed fact is missing, write `{NEEDS: …}` placeholders and flag them in the delivery note — never fabricate a number or quote.
- **Stay in scope.** Write the requested scripts at the requested durations/funnel stages. Don't add extra scripts, change the offer, or redesign the campaign — hand strategy questions back to `/campaign_planner`.
- **No regression.** This upgrade is additive; all original templates, hook formulas, pacing tables, B-roll categories, and emotional arcs remain authoritative.

## Quality Checklist

Before delivering any script:

- [ ] Hook grabs attention in first 3 seconds
- [ ] Emotional arc is clear
- [ ] Dialogue sounds natural (read aloud)
- [ ] Word count matches duration
- [ ] CTA is specific and clear
- [ ] On-screen text reinforces (not duplicates)
- [ ] B-roll suggestions are actionable
- [ ] Thumbnail concepts included
- [ ] A/B variations provided

Attribution

agentik-osagentik-os
View sourceSee grades on GitHubMore from agentik-os →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698461 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →