Design agent skills
Action playbook from three AI Engineer / World's Fair talks. Do not summarize talks — pick a workflow and execute it.
Supporting files (read when needed):
- workflows.md — workflows A–C (steps, deliverables, stop conditions)
- source-index.md — src-NNN → talk learnings in ingest-into-skills
Optional: {SKILL_OUTPUT_DIR}/design-agent-skills/
Step 0 — Pick workflow
Use the decision tree below. Open the matching section in workflows.md.
What is the user trying to do?
├─ Author first skill (structure + triggers) → A
├─ Pair skills with MCP for live data → B
└─ Skill-up coding agents on complex devtools → C
Stop summarizing once a workflow is identified — run its checklist.
Install
cp -r skills/design-agent-skills ~/.claude/skills/
cp -r skills/design-agent-skills ~/.cursor/skills/
cp -r skills/design-agent-skills ~/.codex/skills/
From skills-i-use or ingest-into-skills (playlists/skills-ai-engineer/).
Cross-cutting rules
| Rule | Source |
|---|---|
| SKILL.md slim; workflows in sibling files | [src-002 @ 7:06] |
| Third-person description + triggers | [src-002 @ 35:14] |
| Procedures in skills; state via MCP | [src-001 @ 3:17] |
Disputed steps: see source-index.md. Name workflow A–C; save artifacts to ./skill-outputs/design-agent-skills/ when requested; do not auto-commit.
Invocation examples
@design-agent-skills write a skill for our API
combine MCP with skills for Supabase