agentsclimarketplace

Sb wrappers

Skill strongeron/storybook-workbench/skills/sb-wrappers

AI agent skill bundle for Storybook on React + Vite — audit real-vs-slop components, capture flows, write CSF3 stories, ship. Claude Code / Codex / Cursor.

Install
npx -y skills add strongeron/storybook-workbench --skill sb-wrappers

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 18 stars18 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Scaffold a Storybook-only wrapper — ABCanvas, StateGrid, AppFlowGraph, ProjectInventory, and more. Use for 'compare two designs', 'every state on one canvas', 'the full flow', 'track this decision', or 'scaffold wrappers'.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

5.7 KB, as published. Nobody here has run it

sb-wrappers — composition primitives, not hand-rolled grids

Don't hand-roll grid markup. Scaffold a typed wrapper.

Load only what the task needs (these are two large refs — don't pull both for one wrapper):

  • references/wrapper-library.md — the per-wrapper API. Load it before authoring any wrapper story, to get the props/args of the one you picked.
  • references/composition-patterns.md — the five composition patterns (A/B, role, status grid, page, motion). Load it only for a Compare / multi-pattern / layered build. Do NOT load it for a single StateGrid / ProjectInventory / DesignSystemHealth scaffold — the library ref alone covers that.

Pick the wrapper from intent

User saysWrapperTier
"compare two designs" / "side-by-side"ABCanvas1
"every state on one canvas"StateGrid1
"every role" / "permission UI"StateGrid with role as the varied prop (only if the app is role-gated)1
"track this decision"TrackedDecision + DecisionsDashboard2
"the full flow" / "ordered sequence"StoryStrip / StorySet2
"shader / WebGL / 3D / keyframe motion"ShaderCanvas / R3FCanvas / MotionStage3
"token source of truth" / "DS health" / "inventory"TokensCanvas / DesignSystemHealth / ProjectInventory4
"icon coverage" / "which icons / what sizes / where is an icon used"IconMatrix4
"whole-app route map" / "journey map"AppFlowGraph / JourneyGraphflow

When in the flow (which wrapper, which step)

Wrappers split by what feeds them. The data wrappers render a prior step's JSON — scaffold and use them AFTER that step has run, or they render empty. The component/story wrappers you reach for while writing a story. Suggest the matching wrapper as the next step once its input exists.

Reach for it…Wrapper(s)Needs (run first)
after sb-inventoryProjectInventory, TokenMatrix (Colors foundation — value · mapping · adoption; leave the health prop off).storybook/project-inventory.json
after sb-inventory (usage graph)ComponentUsage, UsageExplorer.storybook/component-pages.json (build-component-pages.py / refresh-usage.sh)
after sb-healthDesignSystemHealth.storybook/design-system-health.json (TokensCanvas auto-discovers — no step)
after sb-flowsAppFlowGraph, JourneyGraph.storybook/flows.json
building Foundations/IconsIconMatrixnothing — scans /src live; pass library + resolve
writing a component story (sb-stories)StateGrid (all states), StateMatrix (variants × states)the component
composing existing storiesStorySet (by tag/id), StoryStrip (ordered)stories already written
iteration (Compare / decide)ABCanvas, TrackedDecision + DecisionsDashboard

Don't scaffold a data wrapper before its step has produced the JSON — scaffold-wrapper.sh prints the generating script for each as a reminder.

States scattered across many stories → consolidate into ONE matrix. When a component's states are spread over separate stories (a story per state — hard to scan, easy to drift), collapse them into a single StateGrid (all states) or StateMatrix (variants × states) so every state is reviewable on one canvas. This is the canonical fix for "the Detection header/list states are scattered — make a concise states view".

Real-usage overlay. StateGrid / StateMatrix take an optional usage prop — usage={usageJson.components.<Name>} (from .storybook/component-usage.json) badges each variant/state with its real call-site count and strikes through declared-but-unused ones (×0 unused). Add it as a flat story in the component's existing CSF, never a new file (a new file makes a new sidebar folder). For colours/type use TokenUsageGrid (reads project-inventory.json tokens.map); for the per-component table, UsageSection embeds into each Docs page via preview.ts docs.page — see sb-inventory.

Scaffold

CORE=${CLAUDE_PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT}}   # scaffold-wrapper is shared core
# scaffold-wrapper.sh is a LOCAL bash script that ships INSIDE this skill — it only copies the bundled
# .tsx wrappers into the project's .storybook/. No network, no install, no external code. See SECURITY.md.
"$CORE/scripts/scaffold-wrapper.sh" --tier 1            # CORE
# --tier 2 | 3 | 4 (adds DESIGN-SYSTEM) | --flow (AppFlowGraph+JourneyGraph) | --all
"$CORE/scripts/scaffold-wrapper.sh" ABCanvas StateGrid  # or specific names

icons.tsx is always copied alongside (the shared icon language) and a barrel index.ts is emitted. Wrappers nest: TrackedDecision → ABCanvas → ShaderCanvas → Hero.

View design

All wrappers follow CONTEXT.md §wrapper-view-design — no emoji, the shared Icon set, and an injectable icons prop on the map wrappers. When you add/edit a wrapper, keep that contract.

Next

Compose your Compare/Flow story, then sb-stories gate (validate-stories.sh).

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.