@musashishao/agent-kit 1.8.1 → 1.9.0

package/.agent/rules/REFERENCE.md

@@ -8,7 +8,7 @@ trigger: always_on
 
 ---
 
-## Available Master Agents (8)
+## Available Master Agents (11)
 
 | Agent | Domain & Focus |
 |-------|----------------|
@@ -19,7 +19,19 @@ trigger: always_on
 | `frontend-specialist` | Frontend & Growth (UI/UX + SEO + Edge/Static Deploy) |
 | `mobile-developer` | Mobile Specialist (Cross-platform + Mobile Performance)|
 | `debugger` | Systematic Root Cause Analysis & Bug Fixing |
-| `game-developer` | Specialized Game Logic & Assets & Performance |
+| `game-developer` | Master Game Dev (Unity, Godot, Unreal, Web Games, VR/AR) |
+| `game-narrative-designer` | Story, Dialogue, LLM-powered NPCs |
+| `game-asset-curator` | Asset Pipeline, Procedural Gen, Optimization |
+| `game-qa-agent` | Automated Playtesting, Balance Analysis, Bug Hunting |
+| `marketing-specialist` | Growth marketing and user acquisition expert |
+| `penetration-tester` | Offensive security specialist (Pentest, Red Team) |
+| `ai-architect` | AI System Design and Multi-Agent Orchestration |
+| `cloud-engineer` | Multi-cloud deployment and serverless architecture |
+| `database-architect` | Database schema design and query optimization |
+| `devops-engineer` | CI/CD, Infrastructure as Code, and Automation |
+| `documentation-writer` | Technical documentation and user manuals |
+| `test-engineer` | Automated testing and quality assurance |
+| `performance-optimizer` | Performance profiling and optimization |
 
 ---
 
@@ -46,6 +58,56 @@ trigger: always_on
 | `observability-patterns` | OpenTelemetry and structured logging for AI |
 | `opentelemetry-expert` | Deep-dive OpenTelemetry SDK and tracing |
 | `ai-incident-management` | AI-specific incident response and playbooks |
+| `llm-app-patterns` | RAG pipelines, agent architectures, LLMOps (NEW) |
+| `langgraph` | LangGraph agent graphs, state management (NEW) |
+| `autonomous-agent-patterns` | Agent loops, tools, permissions, sandboxing (NEW) |
+| `agent-memory-systems` | Working, episodic, semantic memory (NEW) |
+| `crewai` | Multi-agent collaboration, role-based crews (NEW) |
+| `langfuse` | LLM observability, tracing, cost tracking (NEW) |
+| `aws-serverless` | Lambda, API Gateway, DynamoDB, SAM (NEW) |
+| `azure-functions` | HTTP/queue/timer triggers, Durable Functions (NEW) |
+| `gcp-cloud-run` | Cloud Run, Cloud Functions, Firestore (NEW) |
+| `firebase` | Auth, Firestore, Security Rules, Functions (NEW) |
+| `supabase-integration` | Postgres, Auth, RLS, Realtime (NEW) |
+| `neon-postgres` | Serverless Postgres, branching, pooling (NEW) |
+| `graphql` | Schema design, DataLoader, Apollo (NEW) |
+| `clerk-auth` | Next.js auth, organizations, webhooks (NEW) |
+| `bullmq-specialist` | Redis queues, job flows, scheduling (NEW) |
+| `3d-web-experience` | Three.js, React Three Fiber, WebGL/WebGPU (NEW) |
+| `bun-development` | Bun runtime, package manager, bundler (NEW) |
+| `agent-evaluation` | AI agent benchmarking, metrics, A/B testing (NEW) |
+| `browser-automation` | Playwright, Puppeteer, scraping, testing (NEW) |
+| `browser-extension-builder` | Manifest V3, content scripts, popup UIs (NEW) |
+| `discord-bot-architect` | Discord.js, slash commands, interactions (NEW) |
+| `agent-tool-builder` | MCP servers, function calling, sandboxing (NEW) |
+| `ai-agents-architect` | Multi-agent systems, orchestration, communication (NEW) |
+| `ai-product` | AI PM, UX patterns, eval-driven dev (NEW) |
+| `ai-wrapper-product` | AI wrappers, cost optimization, moats (NEW) |
+| `micro-saas-launcher` | Lean SaaS, Stripe, distribution, GTM (NEW) |
+| `inngest` | Durable functions, background jobs, workflows (NEW) |
+| `ethical-hacking-methodology` | Pentest phases (Recon, Scanning, Exploitation) (NEW) |
+| `burp-suite-testing` | Web security proxy, repeater, intruder (NEW) |
+| `active-directory-attacks` | AD poisoning, Kerberoasting, BloodHound (NEW) |
+| `aws-penetration-testing` | AWS IAM auditing, S3 leaks, IMDS (NEW) |
+| `cloud-penetration-testing` | Azure/GCP security auditing, tenant isolation (NEW) |
+| `metasploit-framework` | Exploitation, meterpreter, pivoting (NEW) |
+| `linux-privilege-escalation` | SUID, Sudo, Kernel exploits, Cron (NEW) |
+| `idor-testing` | IDOR, BOLA, parameter pollution (NEW) |
+| `broken-authentication` | JWT bypass, MFA bypass, OAuth flaws (NEW) |
+| `api-fuzzing-bug-bounty` | API discovery, fuzzing, bug bounty (NEW) |
+| `analytics-tracking` | Product analytics, event schemas, funnels (NEW) |
+| `ab-test-setup` | Experiment design, split testing, stats (NEW) |
+| `app-store-optimization` | Mobile app visibility and conversion (NEW) |
+| `copywriting` | Persuasive writing, headlines, CTAs (NEW) |
+| `email-sequence` | Automated flows, onboarding, retention (NEW) |
+| `launch-strategy` | Product Hunt, hype, momentum (NEW) |
+| `marketing-ideas` | Viral loops, pSEO, unconventional growth (NEW) |
+| `paid-ads` | Meta, Google, LinkedIn Ads, ROAS (NEW) |
+| `game-development/unity-integration` | Unity 6, ML-Agents, Sentis, C#/DOTS patterns |
+| `game-development/godot-expert` | Godot 4.3+, GDScript, shaders, RL agents |
+| `game-development/npc-ai-integration` | LLM-powered NPCs, dialogue, personality |
+| `game-development/procedural-generation` | Dungeons, terrain, items, WFC |
+| `game-development/webgpu-shading` | WebGPU, WGSL, Three.js, Babylon.js |
 
 ---
 
@@ -88,3 +150,31 @@ trigger: always_on
 | PII scan | `.agent/skills/privacy-preserving-dev/scripts/pii_scanner.py` |
 | OTel validator | `.agent/skills/observability-patterns/scripts/otel_validator.py` |
 | Trace analyzer | `.agent/skills/opentelemetry-expert/scripts/trace_analyzer.py` |
+
+---
+
+## Workflows (21)
+
+| Command | Purpose |
+|---------|---------|
+| `/ai-agent` | Build AI agent workflow (NEW) |
+| `/marketing` | Marketing campaign planning (NEW) |
+| `/pentest` | Penetration testing workflow (NEW) |
+| `/saas` | SaaS launch workflow (NEW) |
+| `/autofix` | Autonomous self-healing loop |
+| `/brainstorm` | Structured brainstorming |
+| `/context` | Optimal context generation |
+| `/create` | Create new applications |
+| `/dashboard` | Visual project dashboards |
+| `/debug` | Systematic investigation |
+| `/deploy` | Production deployment |
+| `/enhance` | Add/update features |
+| `/next` | Next development steps |
+| `/orchestrate` | Multi-agent coordination |
+| `/plan` | Project planning |
+| `/preview` | Dev server management |
+| `/quality` | Context/output optimization |
+| `/spec` | Specification documents |
+| `/status` | Project/agent status |
+| `/test` | Create and run tests |
+| `/ui-ux-pro-max` | Professional UI/UX planning |

package/.agent/scripts/ak_cli.py

@@ -699,7 +699,7 @@ Examples:
     
     # memory command
     memory_parser = subparsers.add_parser("memory", help="Manage JSON memory")
-    memory_parser.add_argument("mem_command", choices=["init", "get", "set", "update-task", "set-level"])
+    memory_parser.add_argument("mem_command", choices=["init", "get", "set", "update-task", "set-level", "set-lang"])
     memory_parser.add_argument("--type", choices=["brain", "session", "user"], default="session")
     memory_parser.add_argument("--key", help="Key to get or set")
     memory_parser.add_argument("--value", help="Value to set")

package/.agent/scripts/localize_workflows.py

@@ -0,0 +1,54 @@
+import os
+import re
+from pathlib import Path
+
+def localize_workflows(lang, project_root):
+    workflow_dir = Path(project_root) / ".agent" / "workflows"
+    if not workflow_dir.exists():
+        return
+
+    for wf_file in workflow_dir.glob("*.md"):
+        with open(wf_file, "r", encoding="utf-8") as f:
+            content = f.read()
+
+        # Extract frontmatter
+        match = re.match(r"^---\n(.*?)\n---", content, re.DOTALL)
+        if not match:
+            continue
+        
+        frontmatter = match.group(1)
+        remaining = content[match.end():]
+
+        # Parse fields
+        fields = {}
+        for line in frontmatter.split("\n"):
+            if ":" in line:
+                key, val = line.split(":", 1)
+                fields[key.strip()] = val.strip()
+
+        # Backup current description to description_en if not exists
+        if "description" in fields and "description_en" not in fields:
+            fields["description_en"] = fields["description"]
+
+        # Update description based on lang
+        target_key = f"description_{lang}"
+        if target_key in fields:
+            fields["description"] = fields[target_key]
+        elif lang == "en" and "description_en" in fields:
+            fields["description"] = fields["description_en"]
+
+        # Reconstruct frontmatter
+        new_frontmatter = "---\n"
+        for k, v in fields.items():
+            new_frontmatter += f"{k}: {v}\n"
+        new_frontmatter += "---"
+
+        new_content = new_frontmatter + remaining
+        
+        with open(wf_file, "w", encoding="utf-8") as f:
+            f.write(new_content)
+
+if __name__ == "__main__":
+    import sys
+    if len(sys.argv) > 2:
+        localize_workflows(sys.argv[1], sys.argv[2])

package/.agent/scripts/memory_manager.py

@@ -83,7 +83,7 @@ class MemoryManager:
 if __name__ == "__main__":
     import argparse
     parser = argparse.ArgumentParser(description="Manage Agent Kit JSON Memory")
-    parser.add_argument("command", choices=["init", "get", "set", "update-task", "set-level"])
+    parser.add_argument("command", choices=["init", "get", "set", "update-task", "set-level", "set-lang"])
     parser.add_argument("--type", choices=["brain", "session", "user"], default="session")
     parser.add_argument("--key", help="Key to get or set")
     parser.add_argument("--value", help="Value to set")
@@ -113,3 +113,26 @@ if __name__ == "__main__":
             user["skill_level"] = args.level
             manager.save("user", user)
             print(f"✅ User skill level set to: {args.level}")
+    elif args.command == "set-lang":
+        if args.value:
+            user = manager.load("user")
+            if "preferences" not in user:
+                user["preferences"] = {}
+            user["preferences"]["language"] = args.value
+            manager.save("user", user)
+            print(f"✅ Language set to: {args.value}")
+            
+            # Localize workflow descriptions in files
+            try:
+                from localize_workflows import localize_workflows
+                localize_workflows(args.value, manager.project_root)
+                print(f"✅ Workflow descriptions updated to: {args.value}")
+            except ImportError:
+                # Add current directory to path if needed
+                import sys
+                sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+                from localize_workflows import localize_workflows
+                localize_workflows(args.value, manager.project_root)
+                print(f"✅ Workflow descriptions updated to: {args.value}")
+            except Exception as e:
+                print(f"⚠️ Failed to localize workflows: {str(e)}")

package/.agent/skills/3d-web-experience/SKILL.md

@@ -0,0 +1,386 @@
+---
+name: 3d-web-experience
+description: "Build immersive 3D web experiences with Three.js, React Three Fiber, WebGL, and WebGPU. Interactive visualizations, product configurators, games, and creative experiences."
+version: "1.0.0"
+---
+
+# 🎮 3D Web Experience
+
+You are a 3D web developer who creates immersive experiences. You understand WebGL, Three.js, React Three Fiber, and the emerging WebGPU standard. You optimize for performance across devices.
+
+---
+
+## When to Use This Skill
+
+- Interactive 3D product configurators
+- Data visualizations in 3D
+- Web-based games
+- Creative portfolios
+- Architectural visualizations
+- Virtual showrooms
+
+---
+
+## Capabilities
+
+- `threejs`
+- `react-three-fiber`
+- `webgl`
+- `webgpu`
+- `3d-models`
+- `shaders`
+- `physics`
+- `post-processing`
+
+---
+
+## 1. React Three Fiber Setup
+
+```bash
+npm install three @react-three/fiber @react-three/drei
+```
+
+### Basic Scene
+
+```tsx
+// components/Scene.tsx
+'use client'
+
+import { Canvas } from '@react-three/fiber'
+import { OrbitControls, Environment, PerspectiveCamera } from '@react-three/drei'
+
+export function Scene() {
+  return (
+    <Canvas shadows dpr={[1, 2]} gl={{ antialias: true }}>
+      <PerspectiveCamera makeDefault position={[0, 2, 5]} fov={50} />
+      <OrbitControls enableDamping dampingFactor={0.05} />
+      
+      {/* Lighting */}
+      <ambientLight intensity={0.5} />
+      <directionalLight 
+        position={[10, 10, 5]} 
+        intensity={1} 
+        castShadow 
+        shadow-mapSize={[2048, 2048]}
+      />
+      
+      {/* Environment */}
+      <Environment preset="city" />
+      
+      {/* Objects */}
+      <mesh castShadow receiveShadow position={[0, 0.5, 0]}>
+        <boxGeometry args={[1, 1, 1]} />
+        <meshStandardMaterial color="hotpink" />
+      </mesh>
+      
+      <mesh receiveShadow rotation={[-Math.PI / 2, 0, 0]} position={[0, 0, 0]}>
+        <planeGeometry args={[10, 10]} />
+        <meshStandardMaterial color="#444" />
+      </mesh>
+    </Canvas>
+  )
+}
+```
+
+### Interactive Model
+
+```tsx
+// components/InteractiveModel.tsx
+import { useRef, useState } from 'react'
+import { useFrame } from '@react-three/fiber'
+import { useGLTF, Html } from '@react-three/drei'
+import * as THREE from 'three'
+
+export function InteractiveModel({ url }: { url: string }) {
+  const { scene } = useGLTF(url)
+  const ref = useRef<THREE.Group>(null)
+  const [hovered, setHovered] = useState(false)
+  const [active, setActive] = useState(false)
+  
+  useFrame((state, delta) => {
+    if (ref.current && hovered) {
+      ref.current.rotation.y += delta * 0.5
+    }
+  })
+  
+  return (
+    <group ref={ref}>
+      <primitive 
+        object={scene.clone()}
+        scale={active ? 1.2 : 1}
+        onPointerOver={() => setHovered(true)}
+        onPointerOut={() => setHovered(false)}
+        onClick={() => setActive(!active)}
+      />
+      
+      {active && (
+        <Html position={[0, 2, 0]} center>
+          <div className="bg-white p-2 rounded shadow">
+            Product Details
+          </div>
+        </Html>
+      )}
+    </group>
+  )
+}
+
+// Preload model
+useGLTF.preload('/models/product.glb')
+```
+
+---
+
+## 2. Product Configurator
+
+```tsx
+// components/ProductConfigurator.tsx
+import { Canvas } from '@react-three/fiber'
+import { OrbitControls, useGLTF, ContactShadows } from '@react-three/drei'
+import { useState } from 'react'
+
+const COLORS = ['#ff0000', '#00ff00', '#0000ff', '#ffff00', '#ff00ff']
+const MATERIALS = ['standard', 'metallic', 'glass']
+
+function ConfigurableProduct({ color, material }) {
+  const { nodes } = useGLTF('/models/product.glb')
+  
+  const getMaterial = () => {
+    switch (material) {
+      case 'metallic':
+        return <meshStandardMaterial color={color} metalness={0.9} roughness={0.1} />
+      case 'glass':
+        return <meshPhysicalMaterial color={color} transmission={0.9} thickness={0.5} />
+      default:
+        return <meshStandardMaterial color={color} />
+    }
+  }
+  
+  return (
+    <mesh geometry={nodes.Product.geometry}>
+      {getMaterial()}
+    </mesh>
+  )
+}
+
+export function ProductConfigurator() {
+  const [color, setColor] = useState(COLORS[0])
+  const [material, setMaterial] = useState(MATERIALS[0])
+  
+  return (
+    <div className="flex h-screen">
+      {/* 3D Canvas */}
+      <div className="flex-1">
+        <Canvas shadows camera={{ position: [0, 2, 5], fov: 50 }}>
+          <ambientLight intensity={0.5} />
+          <spotLight position={[10, 10, 10]} angle={0.15} penumbra={1} castShadow />
+          
+          <ConfigurableProduct color={color} material={material} />
+          <ContactShadows position={[0, -0.5, 0]} opacity={0.5} blur={2} />
+          
+          <OrbitControls />
+        </Canvas>
+      </div>
+      
+      {/* UI Controls */}
+      <div className="w-64 p-4 bg-gray-100">
+        <h3 className="font-bold mb-4">Customize</h3>
+        
+        <div className="mb-4">
+          <label className="block mb-2">Color</label>
+          <div className="flex gap-2">
+            {COLORS.map(c => (
+              <button
+                key={c}
+                className={`w-8 h-8 rounded-full ${color === c ? 'ring-2' : ''}`}
+                style={{ backgroundColor: c }}
+                onClick={() => setColor(c)}
+              />
+            ))}
+          </div>
+        </div>
+        
+        <div className="mb-4">
+          <label className="block mb-2">Material</label>
+          {MATERIALS.map(m => (
+            <button
+              key={m}
+              className={`block w-full p-2 mb-1 ${material === m ? 'bg-blue-500 text-white' : 'bg-white'}`}
+              onClick={() => setMaterial(m)}
+            >
+              {m}
+            </button>
+          ))}
+        </div>
+      </div>
+    </div>
+  )
+}
+```
+
+---
+
+## 3. Performance Optimization
+
+### Instanced Meshes
+
+```tsx
+// For many identical objects
+import { useRef, useMemo } from 'react'
+import { InstancedMesh, Object3D, Color } from 'three'
+
+function Particles({ count = 1000 }) {
+  const meshRef = useRef<InstancedMesh>(null)
+  
+  const { positions, colors } = useMemo(() => {
+    const positions = new Float32Array(count * 3)
+    const colors = new Float32Array(count * 3)
+    
+    for (let i = 0; i < count; i++) {
+      positions[i * 3] = (Math.random() - 0.5) * 10
+      positions[i * 3 + 1] = Math.random() * 10
+      positions[i * 3 + 2] = (Math.random() - 0.5) * 10
+      
+      const color = new Color().setHSL(Math.random(), 0.7, 0.5)
+      colors[i * 3] = color.r
+      colors[i * 3 + 1] = color.g
+      colors[i * 3 + 2] = color.b
+    }
+    
+    return { positions, colors }
+  }, [count])
+  
+  useEffect(() => {
+    if (!meshRef.current) return
+    
+    const dummy = new Object3D()
+    
+    for (let i = 0; i < count; i++) {
+      dummy.position.set(positions[i * 3], positions[i * 3 + 1], positions[i * 3 + 2])
+      dummy.updateMatrix()
+      meshRef.current.setMatrixAt(i, dummy.matrix)
+    }
+    
+    meshRef.current.instanceMatrix.needsUpdate = true
+  }, [count, positions])
+  
+  return (
+    <instancedMesh ref={meshRef} args={[undefined, undefined, count]}>
+      <sphereGeometry args={[0.05, 16, 16]} />
+      <meshStandardMaterial />
+    </instancedMesh>
+  )
+}
+```
+
+### Level of Detail (LOD)
+
+```tsx
+import { useMemo } from 'react'
+import { Detailed } from '@react-three/drei'
+
+function OptimizedModel() {
+  return (
+    <Detailed distances={[0, 10, 25]}>
+      {/* High detail - close */}
+      <mesh>
+        <sphereGeometry args={[1, 64, 64]} />
+        <meshStandardMaterial />
+      </mesh>
+      
+      {/* Medium detail - mid range */}
+      <mesh>
+        <sphereGeometry args={[1, 32, 32]} />
+        <meshStandardMaterial />
+      </mesh>
+      
+      {/* Low detail - far */}
+      <mesh>
+        <sphereGeometry args={[1, 16, 16]} />
+        <meshStandardMaterial />
+      </mesh>
+    </Detailed>
+  )
+}
+```
+
+---
+
+## 4. Shaders
+
+### Custom Shader Material
+
+```tsx
+import { shaderMaterial } from '@react-three/drei'
+import { extend, useFrame } from '@react-three/fiber'
+import { useRef } from 'react'
+
+const WaveShaderMaterial = shaderMaterial(
+  {
+    uTime: 0,
+    uColor: [0.0, 0.5, 1.0],
+  },
+  // Vertex shader
+  `
+    uniform float uTime;
+    varying vec2 vUv;
+    
+    void main() {
+      vUv = uv;
+      vec3 pos = position;
+      pos.z = sin(pos.x * 5.0 + uTime) * 0.1;
+      pos.z += sin(pos.y * 5.0 + uTime) * 0.1;
+      gl_Position = projectionMatrix * modelViewMatrix * vec4(pos, 1.0);
+    }
+  `,
+  // Fragment shader
+  `
+    uniform vec3 uColor;
+    varying vec2 vUv;
+    
+    void main() {
+      gl_FragColor = vec4(uColor * vUv.y, 1.0);
+    }
+  `
+)
+
+extend({ WaveShaderMaterial })
+
+function WavePlane() {
+  const materialRef = useRef()
+  
+  useFrame(({ clock }) => {
+    if (materialRef.current) {
+      materialRef.current.uTime = clock.elapsedTime
+    }
+  })
+  
+  return (
+    <mesh rotation={[-Math.PI / 2, 0, 0]}>
+      <planeGeometry args={[5, 5, 64, 64]} />
+      <waveShaderMaterial ref={materialRef} />
+    </mesh>
+  )
+}
+```
+
+---
+
+## 5. Performance Checklist
+
+| Check | Description |
+|-------|-------------|
+| ✅ Use `dpr={[1, 2]}` | Optimize for device pixel ratio |
+| ✅ Enable `frameloop="demand"` | Only render when needed |
+| ✅ Use instanced meshes | For many identical objects |
+| ✅ Compress textures | Use KTX2, Basis Universal |
+| ✅ Optimize GLTF | Use Draco compression |
+| ✅ Use LOD | Level of detail for distance |
+| ✅ Dispose resources | Clean up on unmount |
+
+---
+
+## Related Skills
+
+- `react-patterns` - React best practices
+- `frontend-design` - UI/UX design
+- `game-development` - Game patterns

package/.agent/skills/DEPENDENCIES.md

@@ -0,0 +1,54 @@
+# 🗺️ Skill Dependency Graph
+
+This map shows how skills within Agent Kit relate to and depend on each other. Arrows indicate that the source skill provides context or functionality used by the target skill.
+
+## 🤖 AI Development Flow
+```mermaid
+graph TD
+    llm-app-patterns --> autonomous-agent-patterns
+    llm-app-patterns --> ai-agents-architect
+    langgraph --> ai-agents-architect
+    langgraph --> autonomous-agent-patterns
+    agent-memory-systems --> autonomous-agent-patterns
+    agent-tool-builder --> autonomous-agent-patterns
+    agent-evaluation --> ai-product
+    langfuse --> agent-evaluation
+```
+
+## ☁️ Cloud & Infrastructure
+```mermaid
+graph TD
+    aws-serverless --> cloud-penetration-testing
+    gcp-cloud-run --> cloud-penetration-testing
+    firebase --> deployment-procedures
+    supabase-integration --> database-design
+    neon-postgres --> database-design
+    clerk-auth --> broken-authentication
+```
+
+## 📈 Growth & Marketing
+```mermaid
+graph TD
+    analytics-tracking --> ab-test-setup
+    analytics-tracking --> launch-strategy
+    copywriting --> launch-strategy
+    copywriting --> email-sequence
+    marketing-ideas --> launch-strategy
+    paid-ads --> analytics-tracking
+```
+
+## 🛡️ Security Audit
+```mermaid
+graph TD
+    ethical-hacking-methodology --> vulnerability-scanner
+    ethical-hacking-methodology --> penetration-tester
+    vulnerability-scanner --> idor-testing
+    burp-suite-testing --> idor-testing
+    burp-suite-testing --> broken-authentication
+    api-fuzzing-bug-bounty --> idor-testing
+```
+
+## 🛠️ Cross-Domain Dependencies
+- `clean-code`: Global standard for all implementation tasks.
+- `deployment-procedures`: Shared by `cloud-engineer` and `backend-specialist`.
+- `ai-product`: Connects `ai-architect` decisions to `marketing-specialist` outcomes.