design-system-setup
Skill4 starsChanged 21 days agoMITView on GitHub
Set up design-system documentation by running design-expert, design-system-docs, and planning execution subtasks.
In this repository45 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- a11y-auditorSkill
- api-contract-reviewSkill
- backend-releaseSkill
- bug-diagnosisSkill
- client-copy-reviewSkill
- codebase-designSkill
- code-reviewSkill
- create-issueSkill
- create-prSkill
- create-taskSkill
- dependency-upgradeSkill
- deploy-evidenceSkill
- design-expertSkill
- design-system-docsSkill
- domain-modelingSkill
- e2e-verifySkill
- evidence-auditSkill
- feature-discoverySkill
- frontend-releaseSkill
- generate-readmeSkill
- handoffSkill
- implement-taskSkill
- incident-responseSkill
- init-projectSkill
- live-docsSkill
- mcp-debuggingSkill
- memory-curatorSkill
- plan-expertSkill
- planning-featuresSkill
- project-memory-initSkill
- project-memory-refreshSkill
- project-profileSkill
- project-releaseSkill
- proof-runnerSkill
- release-checklistSkill
- researchSkill
- security-hardeningSkill
- security-reviewSkill
- tddSkill
- wiki-forgeSkill
- wiki-initSkill
- wiki-querySkill
- wiki-syncSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 2 are distinctive, and 8 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Run the full design-expert skill as defined in its SKILL.md.
- Very specific: line 19
Read AGENTS.md and DESIGN.md if they exist
- Distinctive: line 20
Run the complete file discovery (Steps 2a–2e) and deep extraction (Steps 3a–3g) — do not skip any scan step
- Very specific: line 21
Write or update DESIGN.md at the project root
- Very specific: line 54
If TASK_MODE is Skip, go directly to Finish.
- Very specific: line 62
Before fetching anything, use AskUserQuestion to ask the user for their ClickUp workspace name or ID:
- Very specific: line 74
Use the hierarchy response to find the workspace that matches WORKSPACE_INPUT (by name or ID).
- Very specific: line 107
Run the full plan-expert skill as defined in its SKILL.md, passing:
- Very specific: line 120
Run the full plan-expert skill as defined in its SKILL.md, passing only:
- Distinctive: line 151
Do not suggest further steps or offer to continue.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
design-system-setup
You are a design system orchestrator. You run three skills in sequence within this context. The skills are preloaded — execute their instructions directly, do not spawn subagents.
Phase 1 — Design Documentation (design-expert)
Run the full design-expert skill as defined in its SKILL.md.
This means:
- Read
AGENTS.mdandDESIGN.mdif they exist - Run the complete file discovery (Steps 2a–2e) and deep extraction (Steps 3a–3g) — do not skip any scan step
- Write or update
DESIGN.mdat the project root - Complete the Step 5 confirmation output
Once DESIGN.md has been written or updated, store its full content as DESIGN_MD_CONTENT and proceed to Phase 2. Do not ask the user whether to continue.
Phase 2 — Storybook Audit or Plan (design-system-docs)
Run the full design-system-docs skill as defined in its SKILL.md.
This means:
- Detect whether Storybook is present (Step 2a) and whether Markdown docs exist (Step 2b)
- Branch to Section A (Storybook audit) or Section B (implementation plan) accordingly
- Complete all audit sub-steps or plan phases in full
- Output the full report or plan
Once the skill completes, capture the entire output as PHASE2_OUTPUT. This is the raw material for Phase 3.
Present the Phase 2 output to the user, then use AskUserQuestion with:
- Header: "Plan the work"
- Question: "How would you like to track the tasks for this work?"
- Options:
ClickUp— create a parent ticket and subtasks in ClickUpLocal files— generate a task plan as local Markdown filesSkip— no task tracking, finish here
Capture the choice as TASK_MODE and proceed to Phase 3.
Phase 3 — Task Creation (plan-expert)
If TASK_MODE is Skip, go directly to Finish.
Branch A — ClickUp (TASK_MODE = ClickUp)
3A.1 — Ask for workspace and list
Before fetching anything, use AskUserQuestion to ask the user for their ClickUp workspace name or ID:
- Header: "ClickUp workspace"
- Question: "Which ClickUp workspace should this ticket be created in? Enter the workspace name or ID."
- Accept free-form text. Capture the answer as
WORKSPACE_INPUT.
Then fetch the workspace hierarchy to resolve available spaces and lists:
mcp__clickup__clickup_get_workspace_hierarchy {}Use the hierarchy response to find the workspace that matches WORKSPACE_INPUT (by name or ID). If no match is found, inform the user and ask them to confirm the workspace name or ID before continuing.
Once the workspace is resolved, use AskUserQuestion to ask the user to pick the destination list:
- Header: "ClickUp list"
- Question: "Which list should the parent ticket be created in?"
- Options: build from the resolved workspace hierarchy (Space → Folder → List). Show at most 6 options; include an "Other (type list ID)" option for anything not listed.
Capture the selected value as LIST_ID. Do not proceed to 3A.2 until both WORKSPACE_INPUT and LIST_ID are confirmed.
3A.2 — Create the parent ticket
Determine the ticket title based on the Phase 2 branch:
- Storybook already exists (Section A was run):
"Improve Storybook implementation" - No Storybook found (Section B was run):
"Set up Storybook and design system documentation"
Build the ticket description from PHASE2_OUTPUT:
- Section A: summarize critical issues, improvements, and coverage gaps
- Section B: summarize current documentation state and include the phase overview
mcp__clickup__clickup_create_task {
list_id: "<LIST_ID>",
name: "<ticket title>",
description: "<summary from PHASE2_OUTPUT>"
}Capture PARENT_TICKET_ID and PARENT_TICKET_URL.
3A.3 — Execute plan-expert
Run the full plan-expert skill as defined in its SKILL.md, passing:
--ticket-id <PARENT_TICKET_ID>--description "<PHASE2_OUTPUT>"— pass the full Phase 2 output so plan-expert has complete context
The skill will present the subtask plan for confirmation, wait for approval, then create all subtasks on the parent ticket. Do not skip the confirmation step — plan-expert owns it.
Branch B — Local files (TASK_MODE = Local files)
3B.1 — Execute plan-expert
Run the full plan-expert skill as defined in its SKILL.md, passing only:
--description "<PHASE2_OUTPUT>"— pass the full Phase 2 output as the planning input
The skill will decompose the work into subtasks, present the plan for confirmation, and on approval create local task files for the session. Do not skip the confirmation step — plan-expert owns it.
Finish
Once all phases are complete, output this summary and stop:
## Design System Setup Complete
### Phase 1 — Design Documentation
DESIGN.md was written / updated at the project root.
### Phase 2 — Storybook
[Storybook audit completed — X critical issues, Y improvements identified]
OR
[Storybook implementation plan produced — 5 phases defined]
### Phase 3 — Task Plan
[ClickUp] Parent ticket: <PARENT_TICKET_URL> — Subtasks created: <count>
OR
[Local files] <count> tasks created for this session
OR
[Skipped]Do not suggest further steps or offer to continue.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

