dna — Claude Code command: what it does, how to install
dna extracts visual styles from a web page and ranks them against a reference corpus across 30 design dimensions.
MIT ★ 5,183 updated 2026-10-23
/plugin marketplace add Manavarya09/design-extract
/plugin install designlang@designlangWhat dna does
The command triggers designlang dna on a target address to measure styling attributes across palette, typography, spacing, shapes, and motion. Unlike flat scorecards, it outputs relative rankings against an indexed collection of real design systems.
Outputs arrive as *-dna.json containing raw metrics with neighbour data, and *-dna.md with an evaluation of outlier axes. Distance measures the mean difference across common features. If only a small portion of the vector was measurable, the low coverage reduces comparative reliability, and the report flags that limitation.
Custom reference sets can be assembled using designlang dna-corpus with product and competitor links. If no corpus exists on disk, the prompt offers to index one before evaluating the target.
What it helps with
- map a website styling against an index of design systems
- identify nearest visual neighbours across 30 design attributes
- spot outlier dimensions in layout, colour, and typography
- determine which styling axes to shift to match a target look
Example prompts for the agent
- /dna <url>
- Run /dna <url> using reference corpus ./my-corpus.json
- Generate a reference set with designlang dna-corpus acme.com competitor.com
How to install dna
1Plugin marketplace
/plugin marketplace add Manavarya09/design-extract/plugin install designlang@designlang2Manually: put the files in place
.claude/commands/dna.mdCall it with /dna. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/dna.mdCall it with /prompts:dna. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/dna.prompt.mdManually: put the files in place
.gemini/commands/dna.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about dna
How to install dna in Claude Code?
Run /plugin marketplace add Manavarya09/design-extract, followed by /plugin install designlang@designlang. The command prompt is located at .claude/commands/dna.md.
What does dna do?
It extracts a 30-feature vector from a webpage, calculates its distance to other design systems in a reference corpus, and exports *-dna.json alongside *-dna.md.
How to build a custom reference corpus for dna?
Run designlang dna-corpus passing target domains, for instance designlang dna-corpus acme.com acme.com/pricing competitor.com.
Source text
dna.md · MIT
--- description: Place a design in the measured design space — nearest systems, per-axis percentiles, outliers --- # /dna Extract a site's design language and locate it in designlang's measured design space: a 30-feature vector covering colour, type, space, shape and motion. Unlike `/grade`, which returns an absolute letter, `/dna` is comparative — every number is a rank against a corpus of real design systems. ## Usage ```bash designlang dna <url> designlang dna <url> --corpus ./my-corpus.json ``` Build your own reference frame — your products, your competitors, whatever you want to be measured against: ```bash designlang dna-corpus acme.com acme.com/pricing competitor.com ``` ## What you get - `*-dna.json` — the vector, the raw measurements behind it, neighbours, percentiles - `*-dna.md` — a readable report: where it sits per axis, its nearest design systems, and the features that make it look the way it does ## Reading the output - **Distance** is the mean absolute difference across the features both designs have — `0.14` means the average feature is 14% of its range apart. - **Percentiles** are ranks within the corpus, not judgements. A design far from the middle is unusual, which may be exactly the intent. - **Coverage** tells you how much of the vector was measurable. A comparison built from a third of the features is a weaker claim, and the report says so. ## Instructions 1. Run `designlang dna <url>` for the URL the user names. 2. Read the emitted `*-dna.md`. 3. Summarise: the nearest systems and what separates them, the axes where the design is an outlier, and — if the user is trying to hit a target look — which axes to move and in which direction. 4. If no corpus exists yet, say so and offer to build one with `dna-corpus`.
© Manavarya Singh · MIT · Manavarya09/design-extract