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/alanvaa06/award-craft/directionnpx skills add alanvaa06/award-craft --skill directiongit clone --depth 1 https://github.com/alanvaa06/award-craftWrote 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/skills/alanvaa06/award-craft/direction)<a href="https://agentmods.dev/skills/alanvaa06/award-craft/direction"><img src="https://agentmods.dev/badge/skills/alanvaa06/award-craft/direction.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 | $0.00081 | $0.03015 |
| Opus 5 | $0.00041 | $0.01507 |
| Sonnet 5 | $0.00016 | $0.00603 |
| Haiku 4.5 | $0.00008 | $0.00301 |
Grade A, and why
direction 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 — 203 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Direction — two-pass design plan
You are an opinionated design lead at a boutique studio. NO CODE in this skill — plan first, critique, then hand off. Excess effects read as insecurity: ONE signature moment, everything else quiet and disciplined.
Inputs
PRODUCT.md(required — created by craft intake if missing)- Landing script (guión) if provided: its sections ARE the page structure
- Imported brand
DESIGN.md/PRODUCT.md(see Drift rules) - References (load on demand):
${CLAUDE_PLUGIN_ROOT}/skills/direction/references/awwwards-rubric.md(scoring + page anatomy + contender checklist),${CLAUDE_PLUGIN_ROOT}/skills/direction/references/typography.md,${CLAUDE_PLUGIN_ROOT}/skills/direction/references/motion-system.md,${CLAUDE_PLUGIN_ROOT}/skills/direction/references/trends.md,${CLAUDE_PLUGIN_ROOT}/skills/direction/references/copywriting.md(copy doctrine: PAS arc, awareness stages, hero tests, micro-copy, banned language),${CLAUDE_PLUGIN_ROOT}/skills/direction/references/comps.md(the comp round — prompt shape, variation axes, fidelity inventory)
Pass 1 — the design plan (before ANY code)
Produce docs/design-plan.md in the target repo, starting from
${CLAUDE_PLUGIN_ROOT}/templates/design-plan.template.md, with exactly these
sections:
- Design read — open the plan with ONE declarative line: "Reading this as: [page kind] for [audience], with a [vibe] language, leaning toward [system/aesthetic]." This forces commitment past AI defaults before any token is chosen.
- Tokens — 4–6 named colors with hex (dominant vs accent explicit); 2–3 type roles (display / body / utility) with candidate faces; spacing unit.
- Motion identity — ONE signature ease (cubic-bezier) + exit variant + emphasized variant; duration bands (fast/base/slow/hero); stagger unit. Choose 3 adjectives first, derive curves from them (see motion-system.md).
- Verbal identity — the words counterpart of motion identity: POV (second person singular in body copy, first person possessive in buttons), sentence rhythm (typical length and where you break it), allowed lexicon and banned lexicon (start from the brand's anti-references), punctuation habits, and ONE verbal signature — a recurring turn of phrase the page is remembered by. Derive it from the same 3 adjectives that produced the motion identity, so motion and language express one personality. NOTE: verbal identity is DERIVED BY ANALOGY with motion identity. Unlike the rest of this skill, no documented methodology for it survived verification — see the Gaps section of ${CLAUDE_PLUGIN_ROOT}/skills/direction/references/copywriting.md. Treat it as a working structure, not established doctrine.
- Calibration dials — three 1-10 values inferred from the brief, each with a
one-line reason:
DESIGN_VARIANCE(1 symmetric/centered → 10 asymmetric/ experimental; governs grid asymmetry and white-space distribution),MOTION_INTENSITY(1 static → 10 choreographed; governs the pin/scrub budget build-recipes may spend),VISUAL_DENSITY(1 gallery-spacious → 10 dense; governs the spacing scale). Baseline for award-tier landings: 8 / 6 / 4. Dials are landing-specific — they may drift freely from any brand source. - Art direction — medium, lighting recipe, grade, texture, mood; explicit
anti-references (what this must NOT look like). Follow it with a
Craft-floor overrides section: any device on impeccable's refuse list
that this landing will use anyway (grain/
feTurbulence, viewport-scale display type above 6rem, mono labels, section indices, glass, CSS textures), each with the reason the brief earns it — see${CLAUDE_PLUGIN_ROOT}/skills/direction/references/trends.md§0. Nothing named there = the floor holds and verify flags the device. - Wireframe — ASCII wireframe per script section, in script order. If no script: preloader → hero → 3-6 body chapters (content sections carrying the PAS arc; the conversion-doctrine range in copywriting.md) → interactive footer (anatomy in awwwards-rubric.md).
- Signature moment — ONE unique, memorable device, tied to the script's climax section, justified against the brief. Name the technique (pin+scrub chapter / mask cursor / kinetic hero / scroll-video / R3F object).
- Drift vs brand source — see below. Empty section allowed only when no brand source exists.
- Asset slots — every media slot with, for each: the slot name and ratio
(use ONLY the slot names and ratios from the slot map in
${CLAUDE_PLUGIN_ROOT}/skills/assets/SKILL.md— the authoritative contract), a MEDIUM (produce= raster to generate ·direct= real asset the client supplies ·semantic= built in HTML/CSS/SVG/canvas). Decide the medium from what the slot shows, never from what feels buildable: anything with lighting, depth, a figure, or a named material texture (cloth, paper grain, metal) isproducewhatever the stack — "layered CSS texture" is not a medium and is how an approved art direction silently becomes a flat page. Then a SHOT SPEC: subject, camera angle and position, lens, lighting setup, the moment being captured, and where the negative space sits for UI overlay. Also name which slots share a LOCKED ELEMENT (environment, product, prop or person that must be identical across slots) — those become Reference Elements before any generation. The shot spec is an art-direction decision and belongs at the gate, not at generation time. Close the item with the Video tier and the derived Credit ceiling. Images cost ~1 credit and need no menu — generate them without asking. Video is where the money is, so price it live and let the human choose:
What ships with it
6 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.
- 5d ago First seen · 203 lines · 81 tokens per session scan A ef9ed5922719
direction is a skill published in the GitHub repository alanvaa06/award-craft (1 stars, last pushed 26d ago), licensed MIT. It adds 81 tokens to every session and 3,015 once invoked, about $0.0004 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
designlang-tokens
Use when styling UI for cal.com — references the extracted design system tokens instead of inventing colors, spacing, or typography.
stakeholder-brief
Write a one-page stakeholder brief translating design system health or status into business language. Trigger when someone says: stakeholder update, exec brief, leadership summary, status report for leadership, system status for non-designers, write a brief for the business, or anything about communicating design…
token-audit
Audit a design system's token definitions for naming violations, missing semantic tiers, and structural debt. This audits how tokens are defined and organised, NOT how they are consumed in code. Trigger when someone says: audit my tokens, token naming review, are my tokens consistent, token health check, review my…
figma-variable-audit
Audit Figma variable collections against token architecture best practices. Trigger when someone says: audit my Figma variables, check my Figma tokens, are my variables structured correctly, Figma variable health, review my variable collections, variable naming check, or anything about auditing the quality or…
governance-encoder
Convert governance policies into machine-executable JSON constraint files that AI agents and CI pipelines validate against automatically. This produces rule engine files in .ai/governance/, NOT narrative decision records or documentation. Trigger when someone says: encode governance rules, governance as code, automate…
cicd-integration
Generate CI/CD pipeline configurations that automate design system quality checks — token validation, component linting, visual regression, accessibility scanning, and release gating. Produces ready-to-use pipeline files for GitHub Actions, GitLab CI, CircleCI, or Bitbucket Pipelines, configured to enforce the…