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 skills/iamk77/skill/layoutnpx skills add IamK77/Skill --skill layoutgit clone --depth 1 https://github.com/IamK77/SkillWhat 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 | $0.00228 | $0.04374 |
| Opus 5 | $0.00114 | $0.02187 |
| Sonnet 5 | $0.00046 | $0.00875 |
| Haiku 4.5 | $0.00023 | $0.00437 |
Grade A, and why
layout 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 yesterday.
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 — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
layout
!checklist init ${CLAUDE_SKILL_DIR} --force
An atelier is a craftsman's working studio — the space where decisions about form, proportion,
and composition happen before the piece leaves the bench. layout is the composition lens of the
atelier suite: where the other skills decide what it looks like (color, type, form, motion),
layout decides how it is organized — the grid, the hierarchy, the reading measure, the spatial
rhythm, and the shape of every application-layer state. Its product is a composed surface: an
explicit armature, a quantified spacing system, and an application layer that holds its form when
reality provides the actual data. It runs across gated stages and will not advance past a GATE
until the checklist tool clears it — order enforced, substance yours.
The governing fact: hierarchy and the grid are decisions, not defaults — and the agent makes neither. The agent places elements in a container, sizes them to fit, and calls it done. It does not ask "what does the eye see first?", does not establish a grid before placing elements, does not consider what happens when a name is sixty characters long or when a table cell contains null. What reads as "amateur layout" is almost always the absence of three things: a decided visual hierarchy (everything at the same weight reads as noise), a consistent spatial system (ad-hoc gaps of 13px, 22px, 7px — the eye detects rhythmic incoherence before the mind names it), and a designed application layer (empty states that are blank divs, error states that are unstyled text, forms where the label disappears when you start typing). These are not polish — they are the substance. A layout without a composed hierarchy is not a draft of a good layout; it is a different artifact.
The boundary with surface:keel: keel owns the CSS mechanism — how Grid and Flexbox distribute space, the cascade, the box model, positioning contexts. layout owns the composition judgment — which grid, which hierarchy model, what density register, how the long tail holds. The questions answered here are design questions that require taste; the questions answered in keel are engineering questions that require correctness. Bring a layout decision here; bring a CSS bug to keel. Cross-reference keel for the mechanism when a stage sends you there for implementation.
This is where the agent era bites:
- The agent does not compose. It fills space. Every element gets a plausible size and a plausible gap; the result is a layout where everything competes equally for attention, nothing reads as the primary entry point, and the eye has no path. Hierarchy must be decided — one primary element, named, placed at the optical center, given dominant size or weight or contrast — before the agent implements it.
- The agent does not design the long tail. It designs the happy path: populated data, English text, a name that is twelve characters. The production surface has names that are sixty characters, numbers that are nine digits, fields that are empty, translations that are forty percent longer. Layout must be stress-tested before it ships, not after the first bug report from production.
- The agent does not design states. An empty state gets a blank div or a developer-default "No data found." An error state gets unstyled red text. A loading state gets a spinner that appears after the content-space has already shifted. The application layer — all states, not just the populated one — must be explicitly designed.
- The agent uses magic px everywhere.
padding: 13px,margin-top: 22px,gap: 7px— each value chosen by eye, none derived from a system. The result is rhythmic incoherence that the eye reads as "off" without the viewer knowing why. All spatial values must come from the spacing scale.
What ships with it
7 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.
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.
- yesterday First seen · 137 lines · 228 tokens per session scan A 8c1aec66a07f
layout is a skill published in the GitHub repository IamK77/Skill (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 228 tokens to every session and 4,374 once invoked, about $0.0011 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 skills, from other repositories
claude-md-improver
Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project…
agent-platform-rag-engine-management
Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…
agent-platform-model-registry
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.
offensive-c2-frameworks
Command and Control framework deployment, configuration, and operational tradecraft for red team engagements. Covers Cobalt Strike (malleable C2 profiles, Beacon types HTTP/HTTPS/DNS/SMB, Beacon Object Files for in-memory execution, sleep and jitter tuning, named pipe pivoting), Sliver (implant generation across…
twitter-reader
Read Twitter/X for financial research using opencli (read-only). Use this skill whenever the user wants to read their Twitter feed, search for financial tweets, view bookmarks, look up user profiles, or gather market sentiment from Twitter/X. Triggers include: "check my feed", "search Twitter for", "show my…
quick-recap
Use when adding or following the red/yellow/green final status block convention for agent responses, especially by installing managed AGENTS.md or CLAUDE.md instructions.