⚡ Agentway

The Autonomous Product Design, Architecture & QA Engine for AI Coding Agents

Transform your AI coding agent into an autonomous Product Designer, UX Researcher, and QA Engineer.

🌐 Quick Navigation

⚡ Quick Start • 🎯 Prompt Playbook • 📊 System Telemetry • 🗺️ Skill Directory (47 Skills) • 🔄 End-to-End Workflow • 💻 CLI Commands • 📖 Official Docs • 💖 Sponsorship

🎯 The Paradigm Shift

Most AI coding assistants generate code in a void. When prompted to "build a SaaS dashboard", they jump straight into rendering divs without validating business assumptions, structuring component tokens, accounting for mobile gestures, or verifying accessibility.

Standard Agent:   Prompt ──────────────────────────────────────────► Code (Guesswork & Inconsistent UI)

With Agentway:    Prompt ──► Discovery ──► Tokens ──► Scaffolding ──► Code ──► Autonomous QA Loop ──► Ship Verdict

Agentway equips any agent with 47 modular engineering contracts. By leveraging Progressive Disclosure, instructions are loaded into context only when relevant — maintaining a 0ms runtime footprint with zero token pollution.

📊 System Telemetry & Key Metrics

| Metric | Specification | Realized Impact | | :--- | :--- | :--- | | 📦 Active Skills | 47 Modular Protocols | Full coverage from customer discovery & pricing to Playwright QA | | 🏷️ Domain Coverage | 11 Specialized Pillars | Product, Behavioral, UX, UI, Delivery, Growth, Mobile, QA, Visual, Modern, Diagnostics | | 🧠 Context Overhead | 0 tokens idle | On-demand progressive loading; never bloats active context | | 🖼️ Visual Extraction | 8-Stage Precision Engine | Extracts tokens from UI screenshots with ≥ 95% visual fidelity | | 💡 Behavioral Monetization | Decoy Effect & Loss Aversion | Mathematical pricing tier optimization with annual discount toggles | | 🌊 Kinematic Motion | Harmonic Spring Oscillator | Damped spring physics ($\zeta = c / 2\sqrt{km}$) replacing linear transitions | | 🩺 Project Diagnostics| Token Adherence Index (TAI) | Objective 0–100 codebase health score across 4 weighted dimensions | | 🔍 Drift Prevention | 50+ Pattern Signatures | Detects hardcoded colors, magic spacing, and CSS anti-patterns with auto-fixes | | 🌓 Theming Model | 4-Tier Surface Elevation | Seamless Light/Dark switching without contrast vibration | | ♿ A11y Standard | WCAG 2.2 AA Target | Automated axe-core + keyboard trap tests + VoiceOver guides | | 🛡️ Self-Healing QA | 3-Iteration Deterministic Loop | Auto-fixes verified markup errors; escalates ambiguous issues | | 🌍 Localization | RTL & 40% Text Expansion | CSS Logical Properties + ICU MessageFormat standards |

⚡ Quick Start

1. Instant Run (No Installation Required)

Execute the interactive setup directly in any workspace:

npx @uwayxt/agent-skills init

2. Global Installation (Recommended)

Keep all 47 skills persistently available across all your projects:

npm install -g @uwayxt/agent-skills
agentway init

Choose Global Installation when prompted. Your IDE (Claude Code, Google Antigravity, Cursor) will auto-discover the skill registry on startup.

🎯 Prompt Playbook: How to Drive Your AI Agent

Crucial Rule: To ensure your agent executes the full 9-phase lifecycle rather than leaping straight into writing unstyled code, use explicit prompt orchestration with mandatory gate approvals.

🌟 Quick-Prompt for New Projects (Greenfield)

Paste this into your AI coding assistant at the start of a new project:

We are building a new project from scratch. Use the complete 47-skill Agentway registry 
in strict sequential order following the official 9-phase workflow. Do not skip any 
skill without my permission, and pause for my approval at the end of each phase before proceeding.

Project Context:
- Idea & Problem: [Describe in 1-2 sentences]
- Visual Reference (if any): [Attach image / link]
- Target Platform: [Web / Mobile / Responsive Both]
- Desired Tech Stack: [e.g., Next.js 15 + Tailwind / Vanilla CSS + Node.js]

PHASE 1 — Problem, Strategy & Monetization: product-discovery, product-strategy, business-model-thinking, mvp-scoping, behavioral-pricing-psychology.
[STOP HERE: Provide discovery_synthesis.md, MVP cut-line & pricing-matrix.json. Wait for approval before Phase 2]

PHASE 2 — Architecture & UX: user-flow-mapping, information-architecture, dashboard-scaffolding-contract, responsive-breakpoint-strategy.
[STOP HERE: Provide user flows, sitemap, and layout contracts. Wait for approval before Phase 3]

PHASE 3 — Visual Language, Tokens & Motion: visual-style-extractor, design-tokens, typography-system, dark-mode-theming-system, kinematic-motion-choreography.
[STOP HERE: Output design-tokens.json & spring-tokens.css. Wait for approval before Phase 4]

PHASE 4 — Resilient Components: design-system-builder, component-style-guide, error-boundary-resilience-design, dashboard-layout-patterns, adaptive-component-behavior, touch-gesture-interaction, perceived-performance-loading.
[STOP HERE: Output UI component matrix with 5-state resilience. Wait for approval before Phase 5]

PHASE 5 — Delivery & Security: security-privacy-review, design-to-code-handoff, module-registry-sync, route-integrity-checker.

PHASE 6 — Autonomous QA Loop: prd-traceability-matrix, interactive-element-audit, flow-based-functional-testing, visual-responsive-regression-testing, accessibility-runtime-audit (WCAG 2.2 AA), ux-chaos-monkey, qa-feedback-loop-orchestrator.

PHASE 7 & FINAL GATE — Project Intelligence & Ship Verdict: project-health-diagnostics (Compute TAI score), design-drift-detector, cross-functional-review.
Deliverable: Final Health Report Card (TAI ≥ 85) + Explicit SHIP / NO-SHIP verdict.

Tag each output with "[skill: <skill-name>]". Do not write implementation code until Phases 1–3 are approved.

📖 Need full brownfield prompt templates, troubleshooting guides, and deep examples?
👉 Read the Complete Agentway Prompt Playbook (docs/PROMPT-PLAYBOOK.md)

💻 Interactive Terminal Showcase

$ agentway list

  █████╗  ██████╗ ███████╗███╗   ██╗████████╗██╗    ██╗ █████╗ ██╗   ██╗
 ██╔══██╗██╔════╝ ██╔════╝████╗  ██║╚══██╔══╝██║    ██║██╔══██╗╚██╗ ██╔╝
 ███████║██║  ███╗█████╗  ██╔██╗ ██║   ██║   ██║ █╗ ██║███████║ ╚████╔╝ 
 ██╔══██║██║   ██║██╔══╝  ██║╚██╗██║   ██║   ██║███╗██║██╔══██║  ╚██╔╝  
 ██║  ██║╚██████╔╝███████╗██║ ╚████║   ██║   ╚███╔███╔╝██║  ██║   ██║   

  ✔ 47 Agent Skills Verified & Active across 11 Domains
  ✔ Model Compatibility: Claude 3.7 Sonnet, GPT-4o, Gemini 2.0 Flash/Pro, DeepSeek V3
  ✔ Skills Lock: .agents/skills-lock.json (Integrity Verified)

  [Domain K: Behavioral Psychology & Kinetic Motion]
  └─ behavioral-pricing-psychology .. Decoy Effect, Loss Aversion & Pricing Tiers
  └─ kinematic-motion-choreography .. Harmonic Spring Oscillator Physics & Asymmetric Curves

  [Domain J: Project Intelligence & Drift Prevention]
  └─ project-health-diagnostics ..... Token Adherence Index (TAI) & Health Report Card
  └─ design-drift-detector .......... 50+ regex signatures detecting hardcoded CSS drift

  [Domain H: Visual Intelligence]
  └─ visual-style-extractor ......... Extract design tokens & components from reference images
  └─ cognitive-load-heatmap-prediction First-500ms attention, Fitts/Hick's Law & CFI score
  
  [Domain A: Product Thinking]
  └─ product-discovery .............. Jobs-to-be-Done, actor mapping & assumption tests
  └─ product-strategy ............... RICE / MoSCoW scoring with OKR alignment
  └─ business-model-thinking ........ Unit economics, CAC/LTV & monetization analysis
  └─ mvp-scoping .................... Riskiest assumption isolation & strict scope cut-lines

  [Domain I: Modern Platform Patterns]
  └─ dark-mode-theming-system ....... 4-tier surface elevation & anti-halasi dark tokens
  └─ error-boundary-resilience-design 5-state resilience matrix (empty/offline/degraded)
  └─ security-privacy-review ........ Auth lifecycle, PII masking & GDPR compliance
  └─ ai-feature-ux-patterns ......... Streaming UI, AI citations & human override controls
  └─ i18n-localization-strategy ..... RTL layout logical properties & string tokenization
  
  [Domains B, C, D, E, F, G: UX, UI, Delivery, Growth, Mobile, QA]
  └─ 28 additional engineering skills active...

🗺️ Skill Directory

Harness cognitive behavioral economics and harmonic spring physics for peak conversion and natural UI motion.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | behavioral-pricing-psychology | "pricing table", "decoy effect", "pricing psychology", "paywall UX", "feature gating" | Mathematically models Decoy Effect (asymmetric dominance), Loss Aversion copywriting, and annual billing toggle savings; outputs pricing-matrix.json and PricingTable.jsx. | | kinematic-motion-choreography | "spring animation", "spring physics", "asymmetric motion", "choreography" | Calculates harmonic oscillator spring physics ($\zeta = c / 2\sqrt{km}$) and asymmetric enter/exit choreography; outputs spring-tokens.css and motion-choreography.js. |

Continuous autonomous health auditing and design token drift prevention.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | project-health-diagnostics | "health check", "diagnose project", "TAI score", "agentway doctor" | Computes Token Adherence Index (TAI) score (0–100) across 4 dimensions; outputs health-report.md, health-report.json, and prioritized prescription.md. | | design-drift-detector | "audit:drift", "hardcoded CSS", "token drift", "design drift" | Scans 7 violation categories across 50+ regex patterns; generates drift-report.md, drift-report.json, and find-and-replace drift-fixes.md. |

Eliminate visual style drift by parsing reference screenshots into deterministic token contracts.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | visual-style-extractor | "reference image", "match this design", "copy this style", "extract colors from screenshot" | Generates style-extraction.json mapping colors, font pairings, 4pt/8pt spacing, border-radius, and components with a 25-Point Fidelity Score. | | cognitive-load-heatmap-prediction | "cognitive load", "attention heatmap", "eye tracking", "CTA hierarchy", "decision overload" | Predicts visual attention paths (Z/F-Pattern, Gutenberg), calculates Fitts's Law (ID), Hick's Law (RT), and outputs a Cognitive Friction Index (CFI/100) score. |

Think before you build. Validate before you code.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | product-discovery | "validate my idea", "who is the user", "problem statement" | Generates discovery_synthesis.md, JTBD problem framing, actor dynamics, and validation experiment recipes (smoke test / concierge). | | product-strategy | "prioritize features", "roadmap", "RICE", "MoSCoW" | Generates product_strategy_one_pager.md, RICE mathematical scoring, OKR validation, and explicit "Won't Do" cut-lines. | | business-model-thinking | "how do we make money", "unit economics", "CAC/LTV" | Evaluates monetization mechanics, marginal cost scalability, and freemium conversion thresholds. | | mvp-scoping | "MVP", "smallest version", "cut scope" | Isolates the single riskiest assumption and draws the absolute minimum viable build-line. | | stakeholder-requirement-mapping | "requirements from sales/ops", "conflicting needs" | Bridges sales, finance, operations, and technical constraints into unified priority matrices. | | business-model-reading | "read this business model", "what modules do we need" | Translates a Business Model Canvas directly into modular software functional blocks. |

Consistency is not an aesthetic choice — it is an engineering discipline.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | design-tokens | "design tokens", "color palette", "spacing scale" | Outputs design-tokens.json and CSS Custom Properties for primitive and semantic color, spacing, radius, and elevation layers. | | dark-mode-theming-system | "dark mode", "theme switching", "color scheme", "dark theme" | Establishes 4-tier dark surface elevation (Ground, Base, Elevated, Highest), anti-halasi text contrast, and anti-flicker runtime scripts. | | typography-system | "typography", "font pairing", "type scale" | Defines fluid type scale (clamp()), font pairings, line-height ratios, and semantic text roles. | | design-system-builder | "component library", "UI kit", "build the design system" | Assembles model-agnostic component registries with strict props, variants, and token dependencies. | | component-style-guide | "component states", "hover/disabled states", "state matrix" | Documents full interaction state matrix (Default, Hover, Active, Focus, Disabled, Error) across touch and mouse. | | ai-feature-ux-patterns | "AI features", "prompt UX", "streaming UI", "AI assistant" | Guides streaming response UX, autoscroll pause, thinking animations, AI badges, and non-destructive diff previews. | | error-boundary-resilience-design | "error state", "empty state", "offline mode", "resilient UX" | Defines 5-state resilience matrices (Ideal, First-use Empty, Filtered Empty, Partial Failure, Network/Offline) with clear recovery CTAs. | | micro-interaction-motion-design | "animation", "transitions", "make the UI feel alive" | Sets motion-fast to motion-deliberate duration tokens, custom bezier easing, and mandatory reduced-motion queries. | | dashboard-layout-patterns | "dashboard layout", "sidebar", "admin panel shell" | Selects optimal application shells (sidebar, top-bar, dual-sidebar, bottom-nav, split-panel) and responsive grids. | | visual-hierarchy-review | "visual hierarchy", "layout review", "what should the eye see first" | Audits focal points, Gestalt spacing rhythm, contrast ratios, and single-primary-CTA discipline. |

Responsive means distinct interaction patterns, not merely squished desktop layouts.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | user-flow-mapping | "user flow", "journey map", "how does the user get from A to B" | Maps decision trees, happy paths, edge cases, and graceful error recovery loops. | | information-architecture | "sitemap", "navigation structure", "organize features" | Builds shallow sitemaps (≤3 levels, ≤7 items), RBAC navigation visibility overlays, and ⌘K Command Palette schemas. | | responsive-breakpoint-strategy | "breakpoints", "responsive design", "fluid grid", "container queries" | Defines the breakpoint system, fluid grid strategy, and container-query patterns for cross-device layouts. | | i18n-localization-strategy | "i18n", "localization", "RTL", "multi-language" | Architects string namespaces, CSS Logical Properties for seamless RTL flipping, and 40% text expansion tolerance. | | adaptive-component-behavior | "adapt for mobile", "sidebar to bottom-nav", "component transformation" | Forces pattern transformations per device (tables -> card stacks, sidebars -> bottom navigation). | | touch-gesture-interaction | "touch targets", "swipe gesture", "thumb reachability" | Enforces ≥ 44×44pt touch zones, thumb-reach layout ergonomics, and safe-area insets (env(safe-area-inset-*)). | | perceived-performance-loading | "skeleton loader", "lazy loading", "make it feel fast" | Designs priority-driven skeleton loaders, progressive reveals, and optimistic mutation feedback. | | user-research-synthesis | "synthesize interviews", "what did users say" | Triangulates qualitative observations with quantitative telemetry; ranks themes by Frequency × Severity. | | usability-heuristics-check | "is this UI usable", "heuristic evaluation" | Audits interfaces against Nielsen's 10 usability heuristics with severity-ranked remediation tasks. | | accessibility-review | "accessibility", "a11y", "WCAG", "contrast" | Evaluates color contrast, cognitive load, semantic tags, and accessible headings. |

Build for what the product will become, not just what it is on day one.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | design-to-code-handoff | "implement this design", "handoff to dev", "turn spec into code" | Maps design tokens and state matrices into sequenced, implementation-ready engineering tasks. | | security-privacy-review | "security review", "privacy audit", "data exposure", "GDPR" | Audits auth lifecycles, PII/secret masking by default, URL query safety, and GDPR export/deletion flows. | | dashboard-scaffolding-contract | "design the dashboard shell", "modular dashboard" | Implements manifest-driven shells that dynamically auto-discover new modules without manual nav edits. | | module-registry-sync | "add new module", "register this module", "wire into dashboard" | Generates module.manifest.json contracts and registers routes into navigation registries. | | route-integrity-checker | "broken links", "dead-end buttons", "route audit" | Validates static/dynamic routes in .agents/route-intents.json and specifies branded 404 escape paths. | | growth-impact-review | "adding a new module", "what breaks if I add this" | Evaluates navigation density, cognitive friction, and performance impact after adding new modules. | | cross-functional-review | "is this ready to ship", "final review", "launch readiness" | Runs 4-pass evaluation (Business · UX · UI · Technical) with strict conflict resolution and ship/no-ship verdict. |

Quality is not an afterthought gate. It is an automated self-healing loop.

| Skill | Trigger Phrases | Key Artifacts & Capabilities | | :--- | :--- | :--- | | prd-traceability-matrix | "traceability", "which requirements are implemented", "PRD coverage" | Maintains live requirement-to-route-to-component mapping; tracks implementation completion percentage. | | interactive-element-audit | "buttons without actions", "orphaned elements", "dead links" | Scans every button, link, and form control; auto-generates Playwright audit test suites. | | flow-based-functional-testing | "test this flow", "end-to-end test", "verify the feature works" | Translates user journey steps into Given-When-Then test cases and executable Playwright scripts. | | visual-responsive-regression-testing | "visual regression", "screenshot comparison", "layout broke" | Captures pixel baselines across breakpoints and flags unintended UI regressions. | | accessibility-runtime-audit | "a11y runtime check", "WCAG on real build", "screen reader" | Executes axe-core against live DOM for WCAG 2.2 AA compliance + VoiceOver/NVDA validation. | | responsive-qa-audit | "responsive QA", "mobile audit", "elements overflowing" | Runs automated Playwright viewport tests across 4+ screen sizes to catch horizontal overflow. | | qa-feedback-loop-orchestrator | "orchestrate QA", "self-healing QA", "close the feedback loop" | Batches QA findings, executes deterministic auto-fixes, retries (max 3×), and escalates edge cases with full context. | | ux-chaos-monkey | "stress test", "chaos test", "text overflow", "extreme data", "network failure", "what if it disconnects" | Injects Text Bombs (German/Russian/Finnish), Data Extremes ($0→$999M, NaN, 120-char names, XSS), and Flaky Network (offline, mid-stream cut, 30s timeout) to prove UI resilience. Generates Playwright chaos test suite via agentway chaos:inject. |

🔄 The End-to-End Workflow

                         [User Idea or Reference Image]
                                       │
                                       ▼
                         ┌──────────────────────────┐
                         │  visual-style-extractor  │ (Optional Visual Entry)
                         └─────────────┬────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  1. PROBLEM, STRATEGY & PRICING                                             │
│  product-discovery ──► product-strategy ──► behavioral-pricing ──► MVP      │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  2. ARCHITECTURE & UX                                                       │
│  user-flow-mapping ──► information-architecture ──► dashboard-scaffolding  │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  3. VISUAL LANGUAGE, TOKENS & KINEMATIC MOTION                              │
│  design-tokens ──► typography-system ──► dark-mode ──► kinematic-motion     │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  4. COMPONENTS & RESILIENCE                                                 │
│  design-system-builder ──► ai-feature-ux ──► error-boundary ──► layout      │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  5. DELIVERY & SECURITY                                                     │
│  security-privacy-review ──► design-to-code-handoff                         │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  6. AUTONOMOUS QA FEEDBACK LOOP                                             │
│  prd-traceability ──► interactive-audit ──► a11y-runtime ──► responsive-qa  │
│                                      │                                      │
│                                      ▼                                      │
│                       qa-feedback-loop-orchestrator                         │
│                    (Auto-Fix ──► Retest ──► Escalate)                       │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │
                                       ▼
┌─────────────────────────────────────────────────────────────────────────────┐
│  7. PROJECT INTELLIGENCE & DRIFT PREVENTION                                 │
│  project-health-diagnostics (TAI Score) ──► design-drift-detector           │
└──────────────────────────────────────┬──────────────────────────────────────┘
                                       │ (All Passed)
                                       ▼
                         ┌──────────────────────────┐
                         │ cross-functional-review  │ ──► [PRODUCTION SHIP]
                         └──────────────────────────┘

💻 CLI Reference (10 Automation Tools)

💡 New to Agentway CLI? Read the 📖 Beginner's CLI Automation Suite Guide (docs/CLI-TOOLS-GUIDE.md) for detailed step-by-step examples and outputs for all 10 tools.

agentway <command> [options]

1. Core Lifecycle Commands

| Command | When to run this? | Description | | :--- | :--- | :--- | | agentway init | First time setup | Interactive installer — provisions skills locally (.agents/skills/) or globally. | | agentway list | Skill exploration | Displays all 47 skills categorized by domain with active trigger phrases. | | agentway update | Version upgrade | Checks for skill registry updates and verifies skills-lock.json. | | agentway help | Quick syntax check | Displays command options and runtime arguments. |

2. Project Health & Token Drift Prevention

| Command | When to run this? | Description | | :--- | :--- | :--- | | agentway doctor [dir] | Codebase audit | Computes Token Adherence Index (TAI) (0–100), outputs health-report.md & prescription.md. | | agentway audit:drift [dir] | PR quality check | Scans CSS for 50+ hardcoded anti-patterns; outputs drift-report.md & drift-fixes.md. |

3. Scaffolding & Design System Engineering

| Command | When to run this? | Description | | :--- | :--- | :--- | | agentway tokens:build [file] | Start of project | Compiles JSON tokens into Vanilla CSS Variables, Tailwind v4, and TypeScript types. | | agentway scaffold:module <name> | New screen / slice | Generates a React component slice with built-in 5-state resilience & manifest contract. | | agentway gen:pricing [dir] | (v1.7.0) Paywall / pricing | Generates pricing-matrix.json, interactive PricingTable.jsx (with Decoy Effect), and pricing.css. | | agentway gen:spring [dir] | (v1.7.0) Micro-interactions | Scaffolds harmonic oscillator spring tokens (spring-tokens.css) and physics solver. |

4. Automated Audits & QA Test Generators

| Command | When to run this? | Description | | :--- | :--- | :--- | | agentway audit:routes [dir] | Before shipping | Audits codebase for dead links (href="#"), orphaned buttons without handlers, and missing aria-label. | | agentway audit:cognitive [dir] | UX review | Calculates Hick's Law decision time & Fitts's Law motor effort to produce a CFI/100 score. | | agentway chaos:inject [dir] | Stress testing | Generates Playwright chaos tests: Text Explosions, Data Extremes ($0→$999M), and Flaky Network. | | agentway gen:playwright [dir] | QA test suite | Scaffolds automated Playwright test suites for WCAG 2.2 AA and responsive viewports. |

📖 Official Documentation

• 🎯 Prompt Playbook & Tutorial — Comprehensive guide & copy-paste prompt templates to orchestrate all 47 skills.
• 🛠️ CLI Automation Suite Guide — Beginner's guide with scenarios & visual walkthroughs for all 10 CLI tools.
• 📖 Skill Authoring Guide — Architectural standards and token budget benchmarks for authoring skills.
• 🗺️ Complete Skill Map — Comprehensive dependency flow & trigger catalog across all 47 skills.
• 🍳 Practical Recipes — 11 production recipes (SaaS from Scratch, Dark Mode, AI Streaming UX, RTL).
• 📜 Version Changelog — Release notes and historical upgrade timeline.
• 🤝 Contribution Guidelines — Corporate standards, branching flow, and Conventional Commits.
• 🛡️ Security Policy — Vulnerability reporting, supported versions, and responsible disclosure SLA.
• 📜 Code of Conduct — Community covenant and ethical participation standards.
• 🚀 Release Playbook — SemVer guidelines, tagging protocol, and NPM publishing procedure.

💖 Sponsorship & Support

Agentway is an independently maintained, open-source engineering initiative. If Agentway saves your team time, improves your front-end architecture, or superpowers your AI coding workflows, please consider supporting the project:

Saweria Donation:
👉 https://saweria.co/uwayxt

🤝 Contributing

We welcome community contributions, skill proposals, and CLI enhancements! Please review our Contribution Guide and Code of Conduct before submitting a pull request.

1. Fork the repo and create your branch (feat/skill-<name> or fix/<description>).
2. Adhere to the Skill Quality Scoring & Token Budget Standards.
3. Ensure all skills are 100% model-agnostic with zero vendor dependencies.
4. Verify SHA-256 integrity in skills-lock.json.
5. Submit a Pull Request following our PR Template.

📄 License & Legal Attribution

Agentway Skills Registry & CLI Suite: © 2026 Wahyudi (@uwayxt)
Licensed under the MIT License.

Engineered with mathematical precision for autonomous AI coding agents and human product teams.