~ / rigs / mirunamu00 / pigment

mirunamu00/pigment

Agent-ready design systems. Pick one, ask Claude Code, Codex or any MCP client to apply it, and every screen stays on-system.

↗ GitHub ★ 0 MIT updated 19d ago project Claude CodeCodex
share on X
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
CONTEXT TAX · EVERY TURN
~1.6k tokens
Featherweight · median rig: 2.3k · breakdown
GUARDRAILS
1/5
No YOLO mode · details
Quality 18/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit mirunamu00/pigment/.claude ./rig-pigment  # inspect, then merge into .claude/

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ npx degit mirunamu00/pigment/.agents/skills/hyperframes-cli .claude/skills/hyperframes-cli
$ npx degit mirunamu00/pigment/.agents/skills/hyperframes-registry .claude/skills/hyperframes-registry
$ npx degit mirunamu00/pigment/.agents/skills/hyperframes .claude/skills/hyperframes
$ npx degit mirunamu00/pigment/.agents/skills/template-intake .claude/skills/template-intake
$ npx degit mirunamu00/pigment/.agents/skills/template-update .claude/skills/template-update
$ curl -fsSL --create-dirs -o .claude/agents/design-critic.md https://raw.githubusercontent.com/mirunamu00/pigment/main/.claude/agents/design-critic.md
$ curl -fsSL --create-dirs -o .claude/agents/prompt-author.md https://raw.githubusercontent.com/mirunamu00/pigment/main/.claude/agents/prompt-author.md
$ curl -fsSL --create-dirs -o .claude/agents/reference-analyst.md https://raw.githubusercontent.com/mirunamu00/pigment/main/.claude/agents/reference-analyst.md

Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.

{
  "permissions": {
    "deny": [
      "Edit(content/published.json)",
      "Write(content/published.json)"
    ]
  }
}

Skills (5)

Subagents (3)

design-criticIndependently judges how close agent-built demos of a template come to the reference's design quality, as 0–5 scores against the reference plus gaps (with general prompt fixes) and defects (counted on
prompt-authorWrites or revises a Pigment template package (the design prompt) in content/templates/<id>/<version>/ from an approved analysis or from evaluation fixes, until pnpm harness validate passes. Use in the
reference-analystDesign-language auditor. Takes a captured reference site (the whole-site survey and the deep captures) and extracts everything that makes it look, move and feel like itself — theme, components, intera

Permissions

deny (2)
Edit(content/published.json)
Write(content/published.json)
ask (0)
—
allow (8)
Bash(pnpm harness:*)
Bash(pnpm install)
Bash(pnpm build)
Bash(pnpm typecheck)
Bash(pnpm test)
Bash(pnpm check:content)
Bash(pnpm --filter:*)
Bash(npx playwright test:*)

Similar rigs

copied ✓