extract

extract is a command for coding agents from Manavarya09/design-extract. It costs 27 tokens per session (462 once invoked), scanned A, original, MIT.

A command that examines a website and extracts its design language—the repeatable visual and writing rules behind the pages—into files for design and development tools.

In plain words
What is it for?
Extracting design tokens, Tailwind styles, Figma variables, motion details, writing tone, and component patterns from a URL.
Why use it?
It saves developers from manually measuring colours, fonts, spacing, components, and other patterns across a site.

Command

Part of the designlang plugin — 2 skills, 14 commands shipped together

About the project

design-extract is a command-line tool that examines live websites with a headless browser and turns their visual and interaction rules into design-system files. It is used by developers and designers to extract tokens, component details, responsive behavior, accessibility findings, and platform-specific outputs, while the catalogue commands, skills, plugin, and rule expose that process to coding agents.

Manavarya09/design-extract · 4,000 stars · on GitHub

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/manavarya09/design-extract/extract
Clone the repo
git clone --depth 1 https://github.com/Manavarya09/design-extract

Or install designlang, the plugin that ships this one along with the rest of its 2 skills, 14 commands.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for extract

README.md
[![agentmods](https://agentmods.dev/badge/commands/manavarya09/design-extract/extract.svg)](https://agentmods.dev/commands/manavarya09/design-extract/extract)
Your own site
<a href="https://agentmods.dev/commands/manavarya09/design-extract/extract"><img src="https://agentmods.dev/badge/commands/manavarya09/design-extract/extract.svg" alt="Measured on agentmods" height="20"></a>
Per session 27 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 462 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00027 $0.00462
Opus 5 $0.00014 $0.00231
Sonnet 5 $0.00005 $0.00092
Haiku 4.5 $0.00003 $0.00046

Measured 5d ago against content hash d144a3ad6df8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

extract scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/extract.md · 50 lines

What it actually says

Run designlang against the user-provided URL and surface the result.

npx designlang $ARGUMENTS

If no $ARGUMENTS were supplied, ask the user which URL to extract.

Default output goes to ./design-extract-output/. Once it finishes, read the generated *-design-language.md (the AI-optimized markdown) and present a tight summary to the user:

  • Primary palette (hex codes)
  • Type families + scale
  • Spacing base + scale
  • WCAG accessibility score
  • Component patterns detected
  • Notable signals (motion feel, material language, brand voice tone)

Then offer follow-ups:

  • /grade <url> — shareable HTML Design Report Card + SVG badge
  • /battle <urlA> <urlB> — head-to-head graded comparison
  • /remix <url> --as <vocab> — restyle in another vocabulary
  • /pack <url> — bundle every output into one design-system directory

Useful flags the user may pass via $ARGUMENTS:

Flag Effect
--full screenshots + responsive + interactions + deep-interact
--depth <n> crawl N additional canonical pages
--dark also extract dark mode
--platforms ios,android,flutter,wordpress multi-platform emitters
--smart LLM fallback for low-confidence classifiers (supports ATLASCLOUD_API_KEY, OPENAI_API_KEY, or ANTHROPIC_API_KEY)
--cookie-file ./session.json authenticated extraction

Full reference: https://github.com/Manavarya09/design-extract#full-cli-reference

Atlas Cloud quick setup for --smart:

export ATLASCLOUD_API_KEY="<atlascloud-api-key>"
export ATLASCLOUD_MODEL="deepseek-ai/deepseek-v4-pro"
npx designlang https://stripe.com --smart
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 50 lines · 27 tokens per session scan A d144a3ad6df8

Subscribe to this mod's changes

extract is a command published in the GitHub repository Manavarya09/design-extract (4,000 stars, last pushed 4d ago), licensed MIT. It adds 27 tokens to every session and 462 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.