ux-evolve

ux-evolve is a command for coding agents from Laith0003/ux-skill. It costs 117 tokens per session (1,358 once invoked), scanned A, original, MIT.

An automatic improvement loop for a newly generated interface or other design file. It checks the result, applies predefined cleanup passes, and repeats until the score improves enough, stops improving, or reaches five rounds.

In plain words
What is it for?
Use it to refine a generated page or interface before delivery. It is intended for broad polishing, not a single targeted change or an artifact that does not exist yet.
Why use it?
It removes the need to run checks and small polish changes by hand over and over. A score shows when the result has reached the command’s quality target.

Command

Part of the ux-skill plugin — 25 commands, 5 agents shipped together

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/laith0003/ux-skill/ux-evolve
Clone the repo
git clone --depth 1 https://github.com/Laith0003/ux-skill

Or install ux-skill, the plugin that ships this one along with the rest of its 25 commands, 5 agents.

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 ux-evolve

README.md
[![agentmods](https://agentmods.dev/badge/commands/laith0003/ux-skill/ux-evolve.svg)](https://agentmods.dev/commands/laith0003/ux-skill/ux-evolve)
Your own site
<a href="https://agentmods.dev/commands/laith0003/ux-skill/ux-evolve"><img src="https://agentmods.dev/badge/commands/laith0003/ux-skill/ux-evolve.svg" alt="Measured on agentmods" height="20"></a>
Per session 117 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,358 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 $0.00117 $0.01358
Opus 5 $0.00059 $0.00679
Sonnet 5 $0.00023 $0.00272
Haiku 4.5 $0.00012 $0.00136

Measured 4d ago against content hash ab6fab67e5a4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ux-evolve 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 4d 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/ux-evolve.md · 106 lines

How it starts

The opening of the file, as written. The whole thing — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/ux-evolve

You are running the /ux-evolve command. Your job is to auto-iterate the polish loop on a target artifact until the evaluation score crosses 90, plateaus, or hits the 5-round safety cap. Offline, deterministic, no LLM calls inside the loop.

When to use

Triggers: "polish this", "evolve this surface", "improve until score 90+", "auto-fix this file", "run the loop on …", "make it ship-ready", "ten/ten this".

Skip when:

  • The artifact is already at 90+ (run uxskill lint <file> --score-only first to confirm).
  • The user wants a single targeted fix (/ux-fix instead).
  • There's no artifact yet (use /ux-design or /ux-system first to generate one).

Behaviour

The loop is a single Python call. Each round:

  1. Run the linter to get the 0-100 quality score
  2. Run the evaluator to score the 7 axes (hierarchy / coherence / spacing / readability / tone / uniqueness + linter)
  3. Apply the 6 deterministic polish passes
  4. Re-evaluate
  5. Decide: target_hit (>=90), plateau (delta < 5), max_rounds (5), or continue

The 6 polish passes are all idempotent — running them twice produces no further change. Strip inline styles, replace generic CTAs, swap placeholder URLs for data URIs, normalize spacing to the 8pt-ish scale, strip Lorem ipsum, normalize font-weight: bold700.

Quality gate

If the final score is < 65, the loop refuses to commit by default and returns stopped_reason: "gate_failed". The user can override with --force to ship anyway. The recommended next move on a gate failure is to regenerate the artifact via /ux-design with different axis hints (e.g., add forbidden: [low-contrast] if the linter is flagging contrast issues).

Process

1. Resolve target

Ask once (or take from $ARGUMENTS): "Path to the artifact (HTML + optional CSS)?"

Default search order if user says "the last one":

  1. .ux/last-design.html + .ux/last-design.css
  2. The most recent file under out/ matching *.html
  3. Fall back to asking explicitly.

Read the full file on GitHub · 106 lines

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. 4d ago First seen · 106 lines · 117 tokens per session scan A ab6fab67e5a4

Subscribe to this mod's changes

ux-evolve is a command published in the GitHub repository Laith0003/ux-skill (65 stars, last pushed 1mo ago), licensed MIT. It adds 117 tokens to every session and 1,358 once invoked, about $0.0006 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.