Script Evaluator logo

Script Evaluator

Community
AnastasiyaW
script-evaluator

Evaluate video scripts and presentations for flatness, tension, and emotional impact. Use when: 'is this script good', 'review script', 'evaluate video', 'why is this boring', 'flatness check', 'script review', 'improve script', 'rate this video'. Scores 6 dimensions (tension, specificity, emotional arc, hook, customer voice, visual variety), identifies specific problems, and suggests concrete fixes with examples. Do NOT use to generate a new script or scene structure from scratch (use video-narrative-arc), to build the product brief (use product-meaning-extractor), or to render/finish the video (use remotion-production-guide or video-post-production); this only critiques an existing script/scene.

Overview

PublisherAnastasiyaW
Repositorycodex-claude-code-config
Skill namescript-evaluator
Stars
212
Forks
32
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by AnastasiyaW on GitHub. Read the source before you install it.

Installation

Install the Script Evaluator AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/AnastasiyaW/codex-claude-code-config/tree/main/skills/video-production/script-evaluator
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/AnastasiyaW/codex-claude-code-config.git /tmp/codex-claude-code-config
mkdir -p .claude/skills
cp -r /tmp/codex-claude-code-config/skills/video-production/script-evaluator .claude/skills/script-evaluator
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Script Evaluator in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Script Evaluator on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Script Evaluator is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Script Evaluator / Flatness Detector

Review a video script, presentation, or rendered scene code and identify exactly WHY it's boring and HOW to fix it. Works on scripts, storyboards, and Remotion/code-based video scenes.

Preserve facts before scoring style

Check claims against the supplied brief and its sources. Keep factual numbers, comparisons, product capabilities, and customer quotations tied to their evidence and limitations. The examples below are illustrative, not reusable product facts. Never invent a statistic, customer, testimonial, or guarantee to raise a score. Without evidence, use a supported qualitative example or flag the exact claim as unverified while continuing the rest of the review. A creative score is a heuristic, not a measured engagement prediction, factual approval, or authority to publish.

Evaluation: 6 Dimensions (score each 1-10)

1. TENSION - Does the viewer feel something?

ScoreDescription
1-3No enemy/problem. Features listed without context. "We do X, Y, Z" tone
4-6Problem mentioned but vague ("saves time"). Some contrast but generic
7-10Specific visceral enemy. Clear before→after. Viewer thinks "that's me!"

Fix: Name the specific, observed pain. Use attributed customer language or clearly identified paraphrases. Show the "before" state without exaggerating it; add a number only when its source supports the claim.

2. SPECIFICITY - Is it concrete or vague?

Red FlagFix
"Saves time"HOW MUCH? "Saves 45 minutes per photo"
"Better quality"WHAT metric? "Original DPI, pixel-to-pixel"
"Innovative solution"WHAT innovation? "Neural retouching with layer output"
"AI-powered"What does the AI DO? "Detects and removes 12 types of artifacts"
"Trusted by thousands"WHO? "Used by 500+ jewelry studios including [name]"

Rule: Prefer a supported concrete example over a vague adjective. Numbers are optional; retain scope and uncertainty rather than forcing numerical specificity.

3. EMOTIONAL ARC - Does it go somewhere?

ScoreDescription
1-3Same energy start to finish. Feels like a slideshow of facts
4-6Some variation but no clear peak/valley
7-10Identifiable beats. At least one DOWN (tension) and one UP (relief). Last scene feels different from first

Fix: Map scenes to an emotional curve. Must have at least one DOWN (problem) and one UP (amazement/relief). The CTA should feel like resolution, not another slide.

4. HOOK STRENGTH - Will they keep watching?

ScoreDescription
1-3Starts with logo, "We are [company]...", or decorative intro with no information
4-6Has information but doesn't create urgency or curiosity
7-10First frame has surprising information. Creates curiosity gap. Uses customer language. MUST see next scene

Fix: The hook is your ad for your ad. Read the first 3 seconds aloud. If it sounds like a corporate intro, try a relevant question, supported fact, attributed customer quote, or clear visual contrast. A stronger hook must not strengthen a claim beyond its evidence.

5. CUSTOMER VOICE - Does it sound human?

ScoreDescription
1-3"Leveraging cutting-edge technology." "Seamless integration." "World-class results."
4-6Reasonable language but still "written by marketing" feeling
7-10Actual customer phrases. Sounds like describing to a friend. Simple, direct, concrete

Fix: Use relevant available customer reviews with source, attribution, date/context, and faithful wording. No fixed quote quota. If no reviews are available, use plain explanatory copy; never fabricate or present inferred language as a testimonial.

6. VISUAL VARIETY - Is it visually dynamic?

ScoreDescription
1-3Same layout every scene. All text, no imagery. Same animation everywhere
4-6Some variation but key moment doesn't stand out visually
7-102+ visual styles. Pacing matches emotion. Most important scene looks DIFFERENT

Fix: Alternate scene types. After text-heavy scenes, do a visual scene. The most important scene should have a unique visual treatment.

Scoring

TENSION:        _/10
SPECIFICITY:    _/10
EMOTIONAL ARC:  _/10
HOOK:           _/10
CUSTOMER VOICE: _/10
VISUAL VARIETY: _/10
────────────────────
TOTAL:          _/60

VERDICT:
  50-60: Strong on this creative rubric; check factual and task acceptance separately
  40-49: Good, minor tweaks needed  
  30-39: Needs work on weakest dimensions
  20-29: Major rewrite - go back to Product Brief
  <20:   Start over with Product Meaning Extractor

5 Common Flatness Patterns

1. "Feature Parade"

Symptom: Scene 1: Feature A. Scene 2: Feature B. Scene 3: Feature C. CTA. Why flat: No narrative tension. It's a list, not a story. Fix: Add a problem scene BEFORE features. Features answer a question - ask the question first.

2. "Logo-First"

Symptom: Opens with 3-5 seconds of logo animation. Why flat: Nobody cares about your logo yet. They care about their problem. Fix: Move logo to the END. Open with the hook.

3. "Generic Superlatives"

Symptom: "The best solution." "Revolutionary technology." "World-class results." Why flat: Every product says this. These words carry zero information. Fix: Replace with supported specifics. A measured speed comparison needs its source and conditions; otherwise explain the actual workflow improvement without a made-up multiplier.

4. "Missing Middle"

Symptom: Good hook, good CTA, flat middle that walks through features. Why flat: No emotional peak. The demo section needs a WOW moment. Fix: Find the single most impressive thing the product does. Build to it. Make it visually different from everything else.

5. "Uniform Energy"

Symptom: Every scene has same pacing, animation speed, text size. Why flat: No rhythm. Like music at one volume. Fix: Vary tempo: fast cuts → slow hero shot → medium features → fast proof → slow CTA.

Usage

After writing a script or coding scenes:

"Evaluate this script/video using the script-evaluator skill.
Score each dimension 1-10 and give specific fixes for anything below 7."

Also useful for comparing versions:

"Evaluate V1 and V2 side by side. Which is stronger and why?"

Gotchas

  • A high total with one dimension at 2 = video still feels broken. Fix the weakest link first.
  • Choose the weakest relevant dimension from the brief and observed evidence; tension is not a universally established top predictor of engagement.
  • Don't optimize "interesting" at the expense of truth or clarity. A confusing or exaggerated script is not an improvement.
  • This evaluator works on: scripts, storyboards, Remotion TSX code, and even finished videos (describe what you see).

Troubleshooting

  • High creative score but unsupported claim: keep the score separate, identify the exact unsupported claim, and revise or remove that claim before calling the content ready for its requested use.
  • No testimonials or numeric proof: do not penalize truthful copy into fabrication; evaluate supported examples and the intended audience instead.

Source basis

  • FTC: advertisement endorsements — a testimonial cannot supply experience or substantiation that does not exist. Applied here as an evidence-preservation rule, not a legal-compliance certification.

Frequently asked questions

What does the Script Evaluator AI skill do?

Evaluate video scripts and presentations for flatness, tension, and emotional impact. Use when: 'is this script good', 'review script', 'evaluate video', 'why is this boring', 'flatness check', 'script review', 'improve script', 'rate this video'. Scores 6 dimensions (tension, specificity, emotional arc, hook, customer voice, visual variety), identifies specific problems, and suggests concrete fixes with examples. Do NOT use to generate a new script or scene structure from scratch (use video-narrative-arc), to build the product brief (use product-meaning-extractor), or to render/finish the...

Why use Script Evaluator on TypingMind?

Because you install it once and use it with any model. Script Evaluator is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Script Evaluator in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/AnastasiyaW/codex-claude-code-config/tree/main/skills/video-production/script-evaluator. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Script Evaluator?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Script Evaluator?

As many as you like. As long as a model supports skills, you can use Script Evaluator with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Script Evaluator AI skill free?

Yes. It is published on GitHub by AnastasiyaW under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇