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.
npx agentmods add commands/app-builders-club/design-builder/improvegit clone --depth 1 https://github.com/app-builders-club/design-builderWrote 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.
[](https://agentmods.dev/commands/app-builders-club/design-builder/improve)<a href="https://agentmods.dev/commands/app-builders-club/design-builder/improve"><img src="https://agentmods.dev/badge/commands/app-builders-club/design-builder/improve.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00024 | $0.03168 |
| Opus 5 | $0.00012 | $0.01584 |
| Sonnet 5 | $0.00005 | $0.00634 |
| Haiku 4.5 | $0.00002 | $0.00317 |
Grade A, and why
improve 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.
How it starts
The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/design-builder:improve — improve an existing design
You take a piece of existing design and make it concretely better. You DON'T produce a P0-P3 audit report (that's /review's job).
Two modes — pick based on the argument:
- Default (mechanical mode). Specific, scope-limited fixes: token swaps, anti-pattern removal, contrast bumps, motion-property fixes, accessibility-fixable subset. Composition is preserved.
--restructuremode. Composition rebuild for one or more weak sections — replaces a generic section with one anchored on a fresh inspiration_pages reference filtered bygood_for_stage. Use when an audit's creative P1 recommendation is "thread X through" / "rebuild Y as Z" / "section feels generic". DO NOT use mechanical mode for these — the documented v2.0 failure is gluing decorative fragments onto a generic section instead of rebuilding it.
Activate the design skill (skills/design/SKILL.md).
Phase 1 — Resolve target
Parse the argument:
- File path or directory (e.g.
src/pages/landing.tsx,src/components/) — read withRead/Glob+Read. - Figma URL (
figma.com/design/...?node-id=...) — callmcp__plugin_figma_figma__get_design_contextandget_screenshot. screenshots(literal) — listdesign/screenshots/, treat the latest screenshots as the target.--restructureflag — if present, jump to the--restructurebranch in Phase 2 instead of the mechanical-fix branch.- Combination: argument may include multiple, separated by spaces.
If no argument: ask the user what to improve. Don't scan the whole repo by default — too noisy.
Read design/design-system.md, design/style-guide.md, and design/content-library.md if they exist. Fixes must use tokens from design-system.md; a11y findings reference rules from style-guide.md; voice & tone deviations reference principles from content-library.md. Do not invent new tokens unless explicitly requested.
Mode selection heuristic. If the user invoked /improve after a /review that flagged composition / distinctiveness / "section feels generic" / "thread X through" issues — default to suggesting --restructure for those specific sections. If user invoked /improve on a fresh design (no prior review) — default to mechanical mode. When ambiguous, ask: "Mechanical patches (tokens, anti-patterns, a11y) or restructure (rebuild a section from a fresh reference)?"
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.
- 5d ago First seen · 152 lines · 24 tokens per session scan A 60a6bdaf32a1
improve is a command published in the GitHub repository app-builders-club/design-builder (4 stars, last pushed 2mo ago), licensed MIT. It adds 24 tokens to every session and 3,168 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 2026-08-31.
Other commands, from other repositories
learn
Force claude-smart to extract learnings from this session now.
init
Install the formatters this repository needs, with every command visible before it runs.
genshijin-compress
Markdown・テキストファイルを原始人形式へ安全圧縮.
standup
Show a daily standup summary with completed, in-progress, and blocked tasks across all active epics.
check-dev
Type-check a Z specification with fuzz.
adr-supersede
Create a new ADR that supersedes an existing one.