{
  "markdown": "# agency-skill\n\n**Agency Designer** is a cross-platform AI skill that brings agency-grade design, agent engineering, and conversion marketing into any AI IDE or coding environment.\n\nInstall it once. Use it in Claude, Copilot, Cursor, Gemini, Goose, Kiro, Windsurf — any environment that supports skills or instruction files.\n\n[![NPM Version](https://img.shields.io/npm/v/agency-skill?color=blue&style=flat-square)](https://www.npmjs.com/package/agency-skill)\n[![License](https://img.shields.io/badge/license-MIT-blue?style=flat-square)](LICENSE)\n\n---\n\n## What It Does\n\nThree commands. Each one turns a vague idea into structured, production-ready output.\n\n| Command | What it gives you |\n|---|---|\n| `/web-design` | A full design system — palette, typography, layout, motion, components — and a `design.md` your team can build from |\n| `/agents-make` | A production-grade agent definition with model, system prompt, tools, MCP servers, skills, and callable agents — Anthropic Managed Agents style |\n| `/marketing-pro` | Conversion-focused copy — headlines that follow the [Result without Pain] formula, specific CTAs, structured page copy, and an audit score |\n\nEvery command interrogates your brief one question at a time, selects the right sub-agents, presents a numbered plan, and waits for your approval before doing anything.\n\n---\n\n## Install\n\n**Bootstrap into any project:**\n\n```bash\nnpx agency-skill init\n```\n\nThis copies the slash commands, agent definitions, and design resources directly into your project.\n\n**Via Skills Registry:**\n\n```bash\nskills install agency-skill\n```\n\n**Via NPM:**\n\n```bash\nnpm install agency-skill\n```\n\n---\n\n## How It Works\n\n1. Type `/web-design`, `/agents-make`, or `/marketing-pro` in your AI IDE\n2. Answer a few brief questions — one at a time, no walls of text\n3. Review the proposed plan\n4. Approve it, then let the skill execute with the right agents in scope\n\n---\n\n## /web-design\n\nProduces a complete design system for any product, page, or component.\n\n**What you get:**\n- Color palette as HSL CSS variables (psychology-first, not random)\n- Premium font pairing with full type scale\n- Layout structure and component language\n- GSAP / Framer Motion animation layer\n- A `design.md` file written to your project root — your living design system doc\n\n**Sub-commands:**\n```\n/web-design component [name]   → generate a single component\n/web-design page [type]        → full page (saas-landing, portfolio, agency, etc.)\n/web-design palette [emotion]  → psychology-first color system\n/web-design font [personality] → premium font pairing\n/web-design animate [type]     → GSAP / Framer / Anime.js animation\n/web-design audit              → quality rubric score\n/web-design style [1-100]      → apply one of 100 graphic design styles\n```\n\n---\n\n## /agents-make\n\nDesigns a real agent system — not a vague description of one.\n\nOutput follows the Anthropic Managed Agents definition format:\n\n```json\n{\n  \"name\": \"Your Agent\",\n  \"model\": \"claude-sonnet-4-6\",\n  \"description\": \"What it does\",\n  \"system\": \"Behavioral rules and constraints\",\n  \"tools\": [{ \"type\": \"agent_toolset_20260401\" }],\n  \"mcp_servers\": [],\n  \"skills\": [],\n  \"callable_agents\": [],\n  \"metadata\": {}\n}\n```\n\nPlus: environment definition, session contract, and quality gates.\n\nWorks for any platform — if you're not on Anthropic Managed Agents, the same schema is adapted to your runtime.\n\n**Sub-commands:**\n```\n/agents-make single     → one production-ready agent definition\n/agents-make pipeline   → sequential pipeline with handoff contracts\n/agents-make director   → director + callable sub-agent tree\n/agents-make managed    → full Anthropic Managed Agents spec\n/agents-make mcp        → MCP server and tool boundary setup\n/agents-make audit      → review an existing system for gaps\n```\n\n---\n\n## /marketing-pro\n\nStructured conversion copy using proven frameworks.\n\n**What you get:**\n- Hero headline in 3 variants (using [Result without Pain] formula)\n- Sub-headline and value proposition\n- Primary and secondary CTAs — never \"Learn More\" or \"Get Started\"\n- Specific social proof copy\n- Section headlines and FAQ objection handlers\n- Conversion audit score\n\n**Frameworks available:** PAS, AIDA, FAB, BAB\n\n**Sub-commands:**\n```\n/marketing-pro headline [context]  → 5 headline variants\n/marketing-pro cta [action]        → CTA copy variants\n/marketing-pro audit               → conversion rubric score\n/marketing-pro rewrite             → full page copy rewrite\n/marketing-pro email [type]        → email sequence\n/marketing-pro seo [keyword]       → SEO-optimised copy\n/marketing-pro ab [element]        → A/B test variants\n```\n\n---\n\n## Supported Environments\n\nWorks in any AI IDE or coding agent that supports skills, instruction files, or slash commands:\n\n- Claude Code\n- Claude Managed Agents\n- GitHub Copilot\n- Cursor\n- Gemini\n- Goose\n- Kiro\n- Windsurf\n- VS Code agent workflows\n- Any custom agent runtime\n\n---\n\n## Agents Included\n\nThe skill ships with five reusable agent role definitions your AI can invoke during any workflow:\n\n- `@visual-director` — palette, typography, and style decisions\n- `@motion-engineer` — GSAP, Framer Motion, and scroll behavior\n- `@copy-strategist` — headlines, CTAs, and conversion copy\n- `@quality-auditor` — S-Tier rubric check (scores 0–50 across typography, color, motion, imagery, copy)\n- `@brief-analyst` — turns vague briefs into structured WHO/WHAT/FEEL/STYLE specs\n\n---\n\n## Anthropic Agent Docs Included\n\nFull Anthropic Managed Agents documentation is bundled so your AI can reference the real spec while designing:\n\n- Agent definition schema\n- Tool configuration (built-in toolset + custom tools)\n- MCP connector setup\n- Skills attachment\n- Environment and session planning\n- Hallucination reduction patterns\n\n---\n\n## License\n\nMIT. Created by [Srinivas Nampalli](https://www.linkedin.com/in/srinivas-nampalli/).\n",
  "bytes": 5856,
  "sha": "02cced94805d86d31900c3b2a891350cdf2548852b1888a739c620b1c9df62e9",
  "repo_slug": "srinivas26k/agency-skill",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_srinivas26k_agency_skill_a5681358/readme"
}