Figma enterprise design-automation workflow: one-screen redesign, greenfield PRD→product pipeline, and progressive memory, with the builder/reviewer/verifier agents.
Builds Figma UI screens with senior-designer craft using the figma-mcp-express tools (live namespace + write path per the orchestrator's brief). Receives a build brief from the figma-redesign or figma-product orchestrator. Component-first, dynamic auto layout, bound tokens only. WRITES to the Figma canvas. Used by /figma-redesign and /figma-product.
Adversarial craft + function verifier for a redesigned or spec-built screen. Owns the judgment dimensions only — D6 craft (PC1–PC13, incl. the hierarchy/squint gate), §1G asset MEANING, §1H functional fidelity, §5F duplicate, D7 pattern adherence. Runs in PARALLEL (background) with figma-structural-verifier, which owns the objective/scannable facts. Neither verifier issues PASS alone — the orchestrator merges both (PASS = structural ∧ craft). READ-ONLY — reports findings, never fixes. Used by /figma-redesign and /figma-product.
Adversarial less-advanced / Sonnet-tier STRUCTURAL verifier for a redesigned or spec-built screen. Independently re-verifies the objective, scannable dimensions — D1 completeness (builds its OWN inventory of the original/spec), D2 component-key fidelity, D3 token binding, D4 auto-layout, D5 content presence, plus the craft SCANS (touch-target ≥44pt, accent-budget ≤2 role-clusters, icon-family consistency, semantic layer names). Runs in PARALLEL (background) with figma-reviewer (craft+function). Neither verifier issues PASS alone — the orchestrator merges both (PASS = structural ∧ craft). READ-ONLY. Used by /figma-redesign and /figma-product.
Dynamic visual researcher for Figma product builds. Finds real product UI/UX reference screenshots, analyzes transferable patterns, and procures concrete media assets with provenance. Writes local reference/asset files and JSON packs only. NEVER mutates Figma.
Manage Figma workflow memory stores. Use when reading or writing project, library, or global playbook memory through the atomic apply script.
Build high-fidelity Figma screens from approved DESIGN.md and COPY.md. Use for final PRD-to-product rendering with component-first, token-bound Figma output.
Create art direction for the Figma product pipeline after PRD research. Use to choose visual direction, concept, references, and craft rules before foundation/spec/build.
Convert an approved direction into a library-bound design foundation. Use to map tokens, components, interaction, brand voice, and platform conventions before specs.
Prepare a PRD for Figma product design. Use to parse requirements, identify screens, flows, states, gaps, audience, and produce a build-ready screen plan.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Build, redesign, and QA real Figma screens with AI coding agents.
Awesome Figma Skills packages a live Figma MCP backend, production design workflows, reusable agent prompts, and implementation-fidelity QA into one public plugin distribution for AI coding agents.
Use it when you want an agent to do more than call raw Figma tools:
figma-mcp-expressRaw Figma MCP tools are powerful, but production design work needs a workflow: bounded reads, component discovery, token binding, visual references, reviewer gates, and safe local artifacts. This repository turns those practices into installable plugins and skills.
The default backend is figma-mcp-express. Development-only server names such
as figma-mcp-express-dev are not required for normal users.
Clone the repository, then run one installer:
git clone https://github.com/Wondermove-Inc/awsome-figima-skills.git
cd awsome-figima-skills
./scripts/install.sh
The installer adds this repository as a Codex plugin marketplace, installs the Figma MCP, Figma automation, and design-QA plugins, and activates the reusable Codex subagents used by those workflows.
Start a new Codex thread after installation so the skills, MCP server config, and agents are loaded.
This repository also includes Claude Code plugin manifests. From Claude Code, add the marketplace and install the same three plugins:
/plugin marketplace add Wondermove-Inc/awsome-figima-skills
/plugin install figma-mcp-express@awesome-figma-skills
/plugin install figma-automation@awesome-figma-skills
/plugin install qa-automation@awesome-figma-skills
For non-interactive setup, Claude Code also exposes equivalent CLI marketplace commands:
claude plugin marketplace add Wondermove-Inc/awsome-figima-skills
claude plugin install figma-mcp-express@awesome-figma-skills
claude plugin install figma-automation@awesome-figma-skills
claude plugin install qa-automation@awesome-figma-skills
Install and run the Figma backend from the Figma MCP Express repository. Keep the Figma Desktop plugin open while an agent is working. Most live Figma workflows require the target file to be open in Figma Desktop with the plugin running.
Figma MCP Express talks directly to the local Figma Desktop file, avoiding official MCP service quotas and keeping high-volume automation responsive.
Agents without Codex or Claude Code plugin support can use this repository as a
skill-file distribution. Import the relevant SKILL.md files into the agent's
own skill or instruction system, keeping each skill's sibling references/,
scripts/, assets/, and agents/ folders with it.
Start with the router skill for broad Figma work, then add the workflow skills you need:
| Workflow | Use it for | What makes it production-oriented |
|---|---|---|
figma-redesign | Rebuild one existing Figma screen into a target library. | Inventory-first rebuild, component-key checks, mechanical pre-gate, two-verifier review. |
figma-product | Turn a PRD into a high-fidelity Figma product. | PRD parsing, direction, foundation, spec, canon screen, gated build fan-out. |
figma-playbook | Load, apply, reflect, and consolidate design memory. | Library/project/global memory stores with proposal-based writes. |
figma-visual-researcher | Fetch current UI references and assets. | Structured request validation, source-year constraints, ranked candidates. |
design-qa | Compare a live app route against Figma. | Screen mapping, Playwright snapshots, bounded Figma reads, structured reports. |
Codex example prompts:
npx claudepluginhub wondermove-inc/awsome-figima-skills --plugin figma-automation개발 워크플로우 자동화 플러그인 - 10개 액티브 스킬 + 8개 패시브 스킬 + 24개 에이전트 + 27개 훅
Design-QA implementation-fidelity review: verify a live rendered app matches its Figma design, with the adversarial design-qa-reviewer agent and a bundled Playwright snapshot script.
Compact Figma MCP with validated batch ops, multi-channel routing, library automation, and design skills.
Harness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
v9.54.1 — Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
AI-powered development tools for code review, research, design, and workflow automation.