How do I install the Source Command Init skill?

Run npx skillmds add tangchunwu/source-command-init in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Source Command Init safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Source Command Init?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Source Command Init free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Source Command Init?

tangchunwu (@tangchunwu) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Source Command Init by tangchunwu — Agent Skill

Name: Source Command Init
Rating: 4.2 (5 reviews)
Author: tangchunwu

File contents

source-command-init

Use this skill when the user asks to run the migrated source command init.

Command Template

Required Reading — Do This First

Before writing any code, read these files completely:

~/.Codex/skills/interface-design/SKILL.md — the foundation, principles, and checks
~/.Codex/skills/interface-design/references/principles.md — detailed craft including subtle layering
~/.Codex/skills/interface-design/references/example.md — how decisions translate to code

Do not skip this. The craft knowledge is in these files.

Scope: Dashboards, apps, tools, admin panels. Not landing pages or marketing sites.

Intent First — Answer Before Building

Before touching code, answer these out loud:

Who is this human? Not "users." Where are they? What's on their mind? A teacher at 7am with coffee is not a developer debugging at midnight.

What must they accomplish? Not "use the dashboard." The verb. Grade submissions. Find the broken deployment. Approve the payment.

What should this feel like? In words that mean something. "Clean" means nothing. Warm like a notebook? Cold like a terminal? Dense like a trading floor?

If you cannot answer these with specifics, stop and ask the user. Do not guess. Do not default.

Before Writing Each Component

State the intent AND the technical approach:

Intent: [who, what they need to do, how it should feel]
Palette: [foundation + accent — and WHY these colors fit the product's world]
Depth: [borders / subtle shadows / layered — and WHY]
Surfaces: [your elevation scale — and WHY this temperature]
Typography: [your typeface choice — and WHY it fits the intent]
Spacing: [your base unit]

Every choice must be explainable. If your answer is "it's common" or "it works" — you haven't chosen. You've defaulted.

The test: If another AI given a similar prompt would produce the same output, you have failed. The interface must emerge from THIS user, THIS problem, THIS intent.

Communication

Be invisible. Don't announce modes or narrate process.

Never say: "I'm in ESTABLISH MODE", "Let me check system.md..."

Instead: Jump into work. State suggestions with reasoning.

Suggest + Ask

Lead with your exploration and recommendation, then confirm:

"Domain: [concepts from this product's world]
Color world: [colors that exist in this domain]
Signature: [one element unique to this product]

Direction: [approach that connects to the above]"

[AskUserQuestion: "Does that direction feel right?"]

Flow

Read the required files above (always — even if system.md exists)
Check if .interface-design/system.md exists
If exists: Apply established patterns from system.md
If not: Assess context, suggest direction, get confirmation, build

The skill files contain the craft principles. system.md contains project-specific decisions. You need both.

After Every Task

Offer to save when you finish building UI:

"Want me to save these patterns to .interface-design/system.md?"

Always offer — new patterns should be captured whether system.md exists or not.

---name: source-command-initdescription: Build UI with craft and consistency. For interface design (dashboards, apps, tools) — not marketing sites.--- # source-command-init Use this skill when the user asks to run the migrated source command `init`. ## Command Template ## Required Reading — Do This First Before writing any code, read these files completely: 1. `~/.Codex/skills/interface-design/SKILL.md` — the foundation, principles, and checks 2. `~/.Codex/skills/interface-design/references/principles.md` — detailed craft including subtle layering 3. `~/.Codex/skills/interface-design/references/example.md` — how decisions translate to code Do not skip this. The craft knowledge is in these files. --- **Scope:** Dashboards, apps, tools, admin panels. Not landing pages or marketing sites. ## Intent First — Answer Before Building Before touching code, answer these out loud: **Who is this human?** Not "users." Where are they? What's on their mind? A teacher at 7am with coffee is not a developer debugging at midnight. **What must they accomplish?** Not "use the dashboard." The verb. Grade submissions. Find the broken deployment. Approve the payment. **What should this feel like?** In words that mean something. "Clean" means nothing. Warm like a notebook? Cold like a terminal? Dense like a trading floor? If you cannot answer these with specifics, stop and ask the user. Do not guess. Do not default. ## Before Writing Each Component State the intent AND the technical approach: ``` Intent: [who, what they need to do, how it should feel] Palette: [foundation + accent — and WHY these colors fit the product's world] Depth: [borders / subtle shadows / layered — and WHY] Surfaces: [your elevation scale — and WHY this temperature] Typography: [your typeface choice — and WHY it fits the intent] Spacing: [your base unit] ``` Every choice must be explainable. If your answer is "it's common" or "it works" — you haven't chosen. You've defaulted. **The test:** If another AI given a similar prompt would produce the same output, you have failed. The interface must emerge from THIS user, THIS problem, THIS intent. ## Communication Be invisible. Don't announce modes or narrate process. **Never say:** "I'm in ESTABLISH MODE", "Let me check system.md..." **Instead:** Jump into work. State suggestions with reasoning. ## Suggest + Ask Lead with your exploration and recommendation, then confirm: ``` "Domain: [concepts from this product's world] Color world: [colors that exist in this domain] Signature: [one element unique to this product] Direction: [approach that connects to the above]" [AskUserQuestion: "Does that direction feel right?"] ``` ## Flow 1. Read the required files above (always — even if system.md exists) 2. Check if `.interface-design/system.md` exists 3. **If exists**: Apply established patterns from system.md 4. **If not**: Assess context, suggest direction, get confirmation, build The skill files contain the craft principles. system.md contains project-specific decisions. You need both. ## After Every Task Offer to save when you finish building UI: "Want me to save these patterns to `.interface-design/system.md`?" Always offer — new patterns should be captured whether system.md exists or not.

Frequently asked questions