vibes-plug 2.14.1 → 3.9.0

package/skills/frontier-ai-models-expert/SKILL.md

@@ -0,0 +1,116 @@
+---
+name: frontier-ai-models-expert
+description: "Expert guide for late-2026 frontier AI models — Claude 5.1 (Fable/Mythos), Project Astra, Gemini 3.1, and Enterprise Frontier Safeguards (EFS) / Panduan ahli model AI frontier akhir-2026 — Claude 5.1, Project Astra, Gemini 3.1, dan Enterprise Frontier Safeguards (EFS)."
+author: "vibes-plug-swarm"
+version: "3.0.0"
+---
+
+# frontier-ai-models-expert — vibes-plug Skill
+
+[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
+
+---
+
+<a name="english"></a>
+## English
+
+### Orchestration & Integration
+Connects and orchestrates with `brainstorming`, `zero-to-prod-orchestrator`, `vercel-ai-sdk-expert`, and `ai-llm-integration-expert` to integrate cutting-edge 2026 frontier models into production applications.
+
+### Description
+Expert guide for integrating late-2026 frontier AI models, focusing on Anthropic's Claude 5.1 (Fable/Mythos) and Google DeepMind's Project Astra (powered by Gemini 3.1). Covers advanced capabilities such as Enterprise Frontier Safeguards (EFS), zero data retention architectures, terminal-bench coding, agentic scientific research, real-time spatial processing, and multimodal tool use.
+
+### Trigger Conditions
+Activate this skill when the user is:
+- Asking to integrate Claude 5.1, Claude Fable, or Claude Mythos.
+- Requesting Project Astra, Gemini 3.1, or universal AI assistant capabilities.
+- Building applications requiring zero data retention or Enterprise Frontier Safeguards (EFS).
+- Developing agentic scientific research pipelines or terminal-based automated coding environments.
+
+---
+
+### Core Concepts
+
+#### 1. Claude 5.1 (Fable & Mythos)
+Claude 5.1 sets a new standard for agentic coding and knowledge work (Terminal-Bench 4.0).
+- **Fable 5.1**: The generally available model optimized for cost (25% less for typical workloads, up to 45% less for agentic work via reduced cache read pricing) and high reasoning capabilities.
+- **Mythos 5.1**: Available through trusted access programs, heavily safeguarded specifically to support high-risk work in cybersecurity (vulnerability discovery, not exploit generation) and life sciences.
+
+#### 2. Enterprise Frontier Safeguards (EFS)
+EFS represents the 2026 standard for data privacy, replacing traditional zero data retention policies. EFS works by storing active data in cloud infrastructure **controlled entirely by the customer**, not the AI provider.
+
+#### 3. Project Astra & Gemini 3.1
+Project Astra explores breakthrough capabilities for universal AI assistants.
+- **Capabilities**: Real-time spatial processing, screensharing, advanced tool use, and multimodal input handling (audio, video, text).
+- **Architecture**: Powered by the Gemini 3.1 architecture, offering unprecedented low-latency interaction and physical world understanding.
+
+---
+
+### Best Practices
+
+1. **Leverage Prompt Caching for Agentic Work**: When using Fable 5.1 for iterative tasks, always utilize prompt caching. The 45% price reduction for cache reads makes it highly efficient for recursive multi-agent reasoning.
+2. **Implement Customer-Controlled Storage for EFS**: For enterprise clients requiring maximum privacy, architecture must support provisioning infrastructure inside the client's VPC to handle EFS compliance.
+3. **Utilize Multimodal Context for Astra**: When integrating Astra-like capabilities, stream video frames and audio concurrently rather than converting everything to text, preserving spatial and temporal context.
+
+---
+
+### Common Pitfalls to Avoid
+
+| Anti-Pattern | Problem | Correct Approach |
+|---|---|---|
+| Relying on provider storage for sensitive data | Fails EFS zero-trust standards | Provision customer-controlled storage buckets for EFS data handling |
+| Using Claude Mythos 5.1 without trusted access | API requests will be rejected | Default to Claude Fable 5.1 for general applications |
+| Converting real-time video to text descriptions | Loses spatial processing capabilities of Astra | Use native multimodal APIs (e.g., Multimodal Live API) to stream raw frames |
+
+---
+
+### Integration with Other Skills (MANDATORY)
+
+This skill works best when combined with:
+- `vercel-ai-sdk-expert` — For implementing Claude 5.1 and Gemini 3.1 streaming and tool calling in React/Next.js.
+- `ai-llm-integration-expert` — For underlying model context protocol (MCP) configurations and RAG setups.
+- `voice-ai-realtime-agent` — For integrating Project Astra's ultra-low latency voice capabilities.
+
+### Referenced By Orchestrators (MANDATORY)
+
+This skill should be referenced by the following orchestrators:
+- `brainstorming` — Add to the "AI & ML" row in the Skill Integration & Orchestration Matrix
+- `zero-to-prod-orchestrator` — Add to Phase 4: Backend & AI Agents
+
+---
+
+<a name="bahasa-indonesia"></a>
+## Bahasa Indonesia
+
+### Integrasi Orkestrasi
+Terhubung dan mengorkestrasi dengan `brainstorming`, `zero-to-prod-orchestrator`, `vercel-ai-sdk-expert`, dan `ai-llm-integration-expert` untuk mengintegrasikan model frontier akhir 2026 ke dalam aplikasi produksi.
+
+### Deskripsi
+Panduan ahli untuk mengintegrasikan model AI frontier akhir 2026, berfokus pada Claude 5.1 (Fable/Mythos) dari Anthropic dan Project Astra dari Google DeepMind (didukung oleh Gemini 3.1). Mencakup kemampuan tingkat lanjut seperti Enterprise Frontier Safeguards (EFS), arsitektur retensi data nol (zero data retention), coding terminal-bench, riset ilmiah agentic, pemrosesan spasial real-time, dan penggunaan tool multimodal.
+
+### Kondisi Pemicu
+Aktifkan skill ini ketika pengguna sedang:
+- Meminta integrasi Claude 5.1, Claude Fable, atau Claude Mythos.
+- Menginginkan Project Astra, Gemini 3.1, atau kapabilitas asisten AI universal.
+- Membangun aplikasi yang memerlukan retensi data nol atau EFS.
+- Mengembangkan pipeline riset ilmiah agentic atau lingkungan coding otomatis berbasis terminal.
+
+### Panduan Singkat
+
+- **Gunakan Fable 5.1 untuk Efisiensi**: Fable 5.1 mengurangi biaya baca cache hingga 45%, sangat ideal untuk pekerjaan agentic yang panjang.
+- **Terapkan EFS untuk Keamanan Enterprise**: Untuk tingkat privasi tertinggi, gunakan Enterprise Frontier Safeguards yang menyimpan data di infrastruktur cloud yang dikendalikan penuh oleh pelanggan.
+- **Manfaatkan Project Astra untuk Multimodal Real-Time**: Gunakan kemampuan Astra untuk pemrosesan spasial, screensharing, dan interaksi latensi rendah tanpa harus mengonversi media menjadi teks terlebih dahulu.
+- **Perhatikan Akses Model**: Gunakan Mythos 5.1 khusus untuk tugas keamanan siber (deteksi kerentanan) dan ilmu hayat melalui program akses terpercaya (trusted access program).
+
+### Integrasi dengan Skill Lain (WAJIB)
+
+Skill ini bekerja paling baik dikombinasikan dengan:
+- `vercel-ai-sdk-expert` — Untuk mengimplementasikan streaming Claude 5.1 dan Gemini 3.1 di React/Next.js.
+- `ai-llm-integration-expert` — Untuk konfigurasi dasar Model Context Protocol (MCP) dan setup RAG.
+- `voice-ai-realtime-agent` — Untuk kapabilitas suara real-time latensi rendah.
+
+### Direferensikan oleh Orchestrator (WAJIB)
+
+Skill ini harus direferensikan oleh orchestrator berikut:
+- `brainstorming` — Tambahkan ke baris "AI & ML" di Matriks Orkestrasi
+- `zero-to-prod-orchestrator` — Tambahkan ke Fase 4 (Backend APIs & AI Agents)

package/skills/fullstack-expert/SKILL.md

@@ -1,184 +1,185 @@
----
-name: fullstack-expert
-description: "Expert-level fullstack development guide covering multi-language (TypeScript, Python, Go, Rust), API design (OpenAPI 3.1, gRPC), microservices, system design patterns, and polyglot architecture / Panduan fullstack tingkat ahli mencakup multi-bahasa, desain API, microservices, dan arsitektur polyglot."
-author: "Roedy Rustam"
----
-
-# Fullstack Expert (2026 Polyglot & Systems Edition)
-
-[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
-
----
-
-<a name="english"></a>
-## English
-
-### Orchestration & Integration
-Connects and orchestrates with domain skills like `brainstorming`, `zero-to-prod-orchestrator`, `senior-frontend`, `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, and `rust-programming-expert`.
-
-### Description
-Expert-level fullstack development across multiple languages and frameworks. Focuses on polyglot backend architectures, system design patterns (Event-Driven Saga, CQRS), API contracts (OpenAPI 3.1, gRPC), and monorepo shared types.
-
-### Trigger Conditions
-- Choosing the right language and framework for a multi-service architecture.
-- Designing API contracts (REST, gRPC, Hono RPC) between heterogeneous services.
-- Implementing distributed system design patterns (Saga, CQRS).
-- Setting up a polyglot monorepo with shared types across frontend and backend.
-
----
-
-### 2026 Polyglot Architecture Matrix
-
-> For comprehensive frontend UI architecture, see `senior-frontend` and `nextjs-app-router-expert`.
-> For dedicated AI agent workflows and LLM integrations, see `ai-llm-integration-expert` and `multi-agent-orchestration`.
-
-| Language | Primary Frameworks | Ideal Workloads |
-|---|---|---|
-| **TypeScript** | Hono, Fastify 5, NestJS | Serverless, Edge APIs, type-safe RPC |
-| **Python** | FastAPI 0.115+, Django 5 | Data science, LLM orchestration, async queues |
-| **Go** | net/http (Go 1.25+), Gin, Echo | Ultra-high throughput microservices, networking, CLIs |
-| **Rust** | Axum 0.8, Actix-web 4 | High-performance compute, memory safety, WASM |
-
-#### Infrastructure & Multi-Service Storage
-- **Primary OLTP**: PostgreSQL with connection pooling (PgBouncer / Supavisor).
-- **Analytics OLAP**: ClickHouse / DuckDB for high-speed aggregations.
-- **Cache & Ephemeral State**: Redis / Upstash with TTL.
-- **Vector Search**: pgvector or Qdrant for semantic search.
-
----
-
-### API Design Standards
-
-#### REST — OpenAPI 3.1
-```yaml
-openapi: "3.1.0"
-info:
-  title: "SaaS Multi-Service API"
-  version: "1.0.0"
-paths:
-  /api/v1/workspaces/{id}:
-    get:
-      summary: "Get workspace by ID"
-      parameters:
-        - name: id
-          in: path
-          required: true
-          schema: { type: string, format: uuid }
-      responses:
-        "200":
-          content:
-            application/json:
-              schema: { $ref: "#/components/schemas/Workspace" }
-        "404":
-          content:
-            application/json:
-              schema: { $ref: "#/components/schemas/ProblemDetail" }
-```
-
-#### gRPC — Protocol Buffers (`proto3`)
-```protobuf
-syntax = "proto3";
-package user.v1;
-
-service UserService {
-  rpc GetUser (GetUserRequest) returns (User);
-  rpc ListUsers (ListUsersRequest) returns (stream User);
-  rpc CreateUser (CreateUserRequest) returns (User);
-}
-
-message User {
-  string id = 1;
-  string email = 2;
-  string name = 3;
-  repeated string workspace_ids = 4;
-}
-```
-
----
-
-### Microservice Patterns
-
-#### Event-Driven (Saga Pattern with Compensations)
-```
-Order Service ──publishes──> "order.created" ──> Payment Service
-                                                      │
-                                               ┌──────┴──────┐
-                                          success?       failure?
-                                               │               │
-                                    "payment.succeeded"  "payment.failed"
-                                               │               │
-                                       Inventory Service  Order Service
-                                       (reserve stock)    (cancel order)
-```
-
-#### CQRS (Command Query Responsibility Segregation)
-- **Write side**: Commands validate and update the primary relational database.
-- **Read side**: Events project into read-optimized denormalized views or ClickHouse tables for high-frequency queries.
-
----
-
-### AI-Native Backend Integration
-```typescript
-import { generateObject } from 'ai';
-import { anthropic } from '@ai-sdk/anthropic';
-import { z } from 'zod';
-
-app.post('/api/analyze', async (c) => {
-  const { content } = await c.req.json();
-  const { object } = await generateObject({
-    model: anthropic('claude-3-7-sonnet-20250219'),
-    schema: z.object({
-      sentiment: z.enum(['positive', 'neutral', 'negative']),
-      topics: z.array(z.string()),
-      summary: z.string().max(200),
-    }),
-    prompt: `Analyze: ${content}`,
-  });
-  return c.json(object);
-});
-```
-
----
-
-### Monorepo with Shared Types
-```typescript
-// packages/types/src/index.ts — single source of truth
-export interface User {
-  id: string;
-  email: string;
-  name: string;
-  plan: 'free' | 'pro' | 'enterprise';
-  isSuperAdmin: boolean;
-}
-```
-
-### Production Security Checklist
-- [ ] User inputs validated with Zod / Pydantic on the server.
-- [ ] Parameterized queries enforced across all database layers.
-- [ ] Rate limiting on all public endpoints (`rate-limit-abuse-prevention`).
-- [ ] Super Admin routes strictly isolated to dedicated subdomain.
-
-## Orchestration & Integration
-- Integrates with: `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, `rust-programming-expert`, `nextjs-app-router-expert`, `api-design-expert`.
-
----
-
-<a name="bahasa-indonesia"></a>
-## Bahasa Indonesia
-
-### Deskripsi
-Panduan pengembangan fullstack tingkat ahli multi-bahasa (TypeScript, Python, Go, Rust), desain kontrak API (OpenAPI 3.1, gRPC), pola arsitektur microservices (Saga, CQRS), dan monorepo shared types.
-
-### Matriks Arsitektur Polyglot 2026
-- **TypeScript**: Hono, Fastify 5, NestJS (Serverless, Cloudflare Edge, Hono RPC).
-- **Python**: FastAPI 0.115+, Django 5 (AI/ML, orkestrasi LLM, pipeline asinkron).
-- **Go**: net/http (Go 1.25+), Gin (Microservices throughput tinggi, CLI).
-- **Rust**: Axum 0.8, Actix-web 4 (Performa ekstrem, komputasi berat, WASM).
-
-### Pola Sistem Terdistribusi
-1. **Saga Pattern**: Kompensasi rollback bertahap saat transaksi multi-layanan mengalami kegagalan.
-2. **CQRS**: Memisahkan database tulis (PostgreSQL) dari view proyeksi baca (ClickHouse/Redis) untuk beban baca tinggi.
-3. **Monorepo Shared Types**: Simpan definisi tipe bersama di `packages/types` sebagai single source of truth antara frontend dan backend.
-
-## Integrasi Orkestrasi
-- Terintegrasi dengan: `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, `rust-programming-expert`, `nextjs-app-router-expert`, `api-design-expert`.
+---
+name: fullstack-expert
+description: "Expert-level fullstack development guide covering multi-language (TypeScript, Python, Go, Rust), API design (OpenAPI 3.1, gRPC), microservices, system design patterns, and polyglot architecture / Panduan fullstack tingkat ahli mencakup multi-bahasa, desain API, microservices, dan arsitektur polyglot."
+author: "Roedy Rustam"
+version: "3.0.0"
+---
+
+# Fullstack Expert (2026 Polyglot & Systems Edition)
+
+[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
+
+---
+
+<a name="english"></a>
+## English
+
+### Orchestration & Integration
+Connects and orchestrates with domain skills like `brainstorming`, `zero-to-prod-orchestrator`, `senior-frontend`, `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, and `rust-programming-expert`.
+
+### Description
+Expert-level fullstack development across multiple languages and frameworks. Focuses on polyglot backend architectures, system design patterns (Event-Driven Saga, CQRS), API contracts (OpenAPI 3.1, gRPC), and monorepo shared types.
+
+### Trigger Conditions
+- Choosing the right language and framework for a multi-service architecture.
+- Designing API contracts (REST, gRPC, Hono RPC) between heterogeneous services.
+- Implementing distributed system design patterns (Saga, CQRS).
+- Setting up a polyglot monorepo with shared types across frontend and backend.
+
+---
+
+### 2026 Polyglot Architecture Matrix
+
+> For comprehensive frontend UI architecture, see `senior-frontend` and `nextjs-app-router-expert`.
+> For dedicated AI agent workflows and LLM integrations, see `ai-llm-integration-expert` and `multi-agent-orchestration`.
+
+| Language | Primary Frameworks | Ideal Workloads |
+|---|---|---|
+| **TypeScript** | Hono, Fastify 5, NestJS | Serverless, Edge APIs, type-safe RPC |
+| **Python** | FastAPI 0.115+, Django 5 | Data science, LLM orchestration, async queues |
+| **Go** | net/http (Go 1.25+), Gin, Echo | Ultra-high throughput microservices, networking, CLIs |
+| **Rust** | Axum 0.8, Actix-web 4 | High-performance compute, memory safety, WASM |
+
+#### Infrastructure & Multi-Service Storage
+- **Primary OLTP**: PostgreSQL with connection pooling (PgBouncer / Supavisor).
+- **Analytics OLAP**: ClickHouse / DuckDB for high-speed aggregations.
+- **Cache & Ephemeral State**: Redis / Upstash with TTL.
+- **Vector Search**: pgvector or Qdrant for semantic search.
+
+---
+
+### API Design Standards
+
+#### REST — OpenAPI 3.1
+```yaml
+openapi: "3.1.0"
+info:
+  title: "SaaS Multi-Service API"
+  version: "1.0.0"
+paths:
+  /api/v1/workspaces/{id}:
+    get:
+      summary: "Get workspace by ID"
+      parameters:
+        - name: id
+          in: path
+          required: true
+          schema: { type: string, format: uuid }
+      responses:
+        "200":
+          content:
+            application/json:
+              schema: { $ref: "#/components/schemas/Workspace" }
+        "404":
+          content:
+            application/json:
+              schema: { $ref: "#/components/schemas/ProblemDetail" }
+```
+
+#### gRPC — Protocol Buffers (`proto3`)
+```protobuf
+syntax = "proto3";
+package user.v1;
+
+service UserService {
+  rpc GetUser (GetUserRequest) returns (User);
+  rpc ListUsers (ListUsersRequest) returns (stream User);
+  rpc CreateUser (CreateUserRequest) returns (User);
+}
+
+message User {
+  string id = 1;
+  string email = 2;
+  string name = 3;
+  repeated string workspace_ids = 4;
+}
+```
+
+---
+
+### Microservice Patterns
+
+#### Event-Driven (Saga Pattern with Compensations)
+```
+Order Service ──publishes──> "order.created" ──> Payment Service
+                                                      │
+                                               ┌──────┴──────┐
+                                          success?       failure?
+                                               │               │
+                                    "payment.succeeded"  "payment.failed"
+                                               │               │
+                                       Inventory Service  Order Service
+                                       (reserve stock)    (cancel order)
+```
+
+#### CQRS (Command Query Responsibility Segregation)
+- **Write side**: Commands validate and update the primary relational database.
+- **Read side**: Events project into read-optimized denormalized views or ClickHouse tables for high-frequency queries.
+
+---
+
+### AI-Native Backend Integration
+```typescript
+import { generateObject } from 'ai';
+import { anthropic } from '@ai-sdk/anthropic';
+import { z } from 'zod';
+
+app.post('/api/analyze', async (c) => {
+  const { content } = await c.req.json();
+  const { object } = await generateObject({
+    model: anthropic('claude-3-7-sonnet-20250219'),
+    schema: z.object({
+      sentiment: z.enum(['positive', 'neutral', 'negative']),
+      topics: z.array(z.string()),
+      summary: z.string().max(200),
+    }),
+    prompt: `Analyze: ${content}`,
+  });
+  return c.json(object);
+});
+```
+
+---
+
+### Monorepo with Shared Types
+```typescript
+// packages/types/src/index.ts — single source of truth
+export interface User {
+  id: string;
+  email: string;
+  name: string;
+  plan: 'free' | 'pro' | 'enterprise';
+  isSuperAdmin: boolean;
+}
+```
+
+### Production Security Checklist
+- [ ] User inputs validated with Zod / Pydantic on the server.
+- [ ] Parameterized queries enforced across all database layers.
+- [ ] Rate limiting on all public endpoints (`rate-limit-abuse-prevention`).
+- [ ] Super Admin routes strictly isolated to dedicated subdomain.
+
+## Orchestration & Integration
+- Integrates with: `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, `rust-programming-expert`, `nextjs-app-router-expert`, `api-design-expert`.
+
+---
+
+<a name="bahasa-indonesia"></a>
+## Bahasa Indonesia
+
+### Deskripsi
+Panduan pengembangan fullstack tingkat ahli multi-bahasa (TypeScript, Python, Go, Rust), desain kontrak API (OpenAPI 3.1, gRPC), pola arsitektur microservices (Saga, CQRS), dan monorepo shared types.
+
+### Matriks Arsitektur Polyglot 2026
+- **TypeScript**: Hono, Fastify 5, NestJS (Serverless, Cloudflare Edge, Hono RPC).
+- **Python**: FastAPI 0.115+, Django 5 (AI/ML, orkestrasi LLM, pipeline asinkron).
+- **Go**: net/http (Go 1.25+), Gin (Microservices throughput tinggi, CLI).
+- **Rust**: Axum 0.8, Actix-web 4 (Performa ekstrem, komputasi berat, WASM).
+
+### Pola Sistem Terdistribusi
+1. **Saga Pattern**: Kompensasi rollback bertahap saat transaksi multi-layanan mengalami kegagalan.
+2. **CQRS**: Memisahkan database tulis (PostgreSQL) dari view proyeksi baca (ClickHouse/Redis) untuk beban baca tinggi.
+3. **Monorepo Shared Types**: Simpan definisi tipe bersama di `packages/types` sebagai single source of truth antara frontend dan backend.
+
+## Integrasi Orkestrasi
+- Terintegrasi dengan: `js-backend-expert`, `go-programming-expert`, `python-programming-expert`, `rust-programming-expert`, `nextjs-app-router-expert`, `api-design-expert`.