A workflow for refining a user interface by checking visual consistency, states, spacing, typography, icons, colours, tooltips, menus, and keyboard shortcuts. It can also check small interface-related backend fixes within the approved scope.

Use it after implementation or during release work to review and refine a web or application interface without automatically redesigning its structure.

A functional interface can still have inconsistent visuals or missing interaction states. This pass gives those details a focused review while requiring approval for structural UI changes.

auto-polish — Skill for Claude Code by Jerry0022/dotclaude

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.

Any agent

npx skills add Jerry0022/dotclaude --skill auto-polish

Clone the repo

git clone --depth 1 https://github.com/Jerry0022/dotclaude

Made for: Claude Code.

Or install devops, the plugin that ships this one along with the rest of its 18 skills, 12 agents, 9 hooks, 1 MCP server.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/jerry0022/dotclaude/auto-polish/github.svg)](https://agentmods.dev/skills/jerry0022/dotclaude/auto-polish)

Your own site

<a href="https://agentmods.dev/skills/jerry0022/dotclaude/auto-polish"><img src="https://agentmods.dev/badge/skills/jerry0022/dotclaude/auto-polish/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/jerry0022/dotclaude/auto-polish"><img src="https://agentmods.dev/badge/skills/jerry0022/dotclaude/auto-polish.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 193 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,138 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00193	$0.03138
Opus 5.5	$0.00077	$0.01255
Sonnet 5.5	$0.00039	$0.00628
Haiku 4.5	$0.00019	$0.00314

Measured 11d ago against content hash f2d77bc661a9, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

auto-polish 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 11d 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.

plugins/devops/skills/auto-polish/SKILL.md · 231 lines

How it starts

The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Refine the UI: consistency, visual polish, state-visuals, and UI-side functionality — with user approval for structural changes. Scope: $ARGUMENTS.

Invocation Context

Same callers as /auto-harden — the user, /do-run and /do-ship, all above this skill in the call graph. /auto-agents (layer 5) never calls it; it is the layer this skill executes through.

Direct — user asks for it (trigger phrase; the skill is hidden from the slash menu).
From /do-run (every run that touched UI files — no longer a question) — --invoked-by=do-run, a full pass scoped to the run's changes, executed through auto-agents (§ Execution); without UI files in scope it self-skips (Step 2). Under "Autonom" do-run adds --autonomous (no prompts; structural changes always flagged, never auto-applied); under "Strikt" it adds --strict. Skip self-spawned qa/redteam when the parent owns those waves. The pre-PR-2 values --invoked-by=agents and --invoked-by=autonomous (the latter implies --autonomous) are read as do-run.
From /do-ship — --invoked-by=ship plus the diff's UI files as scope: the report-only § Rules-only path (ship). /do-ship calls /auto-harden --invoked-by=ship at the same step.
Under strict — --strict (do-run "Strikt", /do-ship under strict mode, or the [claude-strict contract] in context); on the ship path /do-ship then applies none of the mechanical fixes.

Execution — through auto-agents

The changes of a full pass (Steps 5–10) run through /auto-agents, the single execution path: Skill("auto-agents", "--from=auto-polish --mode=<m> <change list>") with --mode=background under --autonomous, else --mode=interactive. Read its result block (tier, done, open, needs-decision, ship) and ignore ship — this skill never ships. Inline shortcut: when your own tier check lands on Inline (one domain, ≤ ~5 files), apply the changes yourself without loading auto-agents. Read-only helpers (scout scans, qa, redteam, designer consults) stay direct Agent spawns. The rules-only path never loads auto-agents.

Read the full file on GitHub · 231 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

11d ago Changed f2d77bc661a9
16d ago Changed · -6 lines ea1b967c5ae9
23d ago Changed 70473a424d2c
1mo ago Changed · -14 lines a1812b387dfe
1mo ago Changed e2d6240d9718
1mo ago Changed 205d23c56d03
1mo ago Changed · -240 lines · +4 tokens per session b6d935b5f26f
1mo ago Changed · +2 lines · +6 tokens per session b997e0f7fb69
1mo ago First seen · 489 lines · 183 tokens per session scan A 639f2c1191b5

Subscribe to this mod's changes

auto-polish is a skill published in the GitHub repository Jerry0022/dotclaude (6 stars, last pushed 11d ago), licensed MIT. It adds 193 tokens to every session and 3,138 once invoked, about $0.0008 per session on Opus 5.5. 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 2026-09-24.

Design and UX

auto-polish next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
auto-polishJerry0022/dotclaude	Skill	6	11d ago	MIT	A	193	Claude Code
webgl-holographic-foilnexu-io/open-design	Skill	106k	yesterday	Apache-2.0	A	41	Claude Code, Codex
html-ppt-hermes-cyber-terminalnexu-io/open-design	Skill	106k	yesterday	Apache-2.0	A	53	Claude Code, Codex
html-ppt-taste-brutalistnexu-io/open-design	Skill	106k	yesterday	Apache-2.0	A	78	Claude Code, Codex
visual-ralphYeachan-Heo/oh-my-codex	Skill	34k	3d ago	MIT	A	52	Codex
glance-canonical-layoutsandroid/skills	Skill	10k	yesterday	Apache-2.0	A	56	Claude Code
accessibilitymicrosoft/hve-core	Skill	1.6k	yesterday	MIT	A	47	Claude Code

auto-polish