# Pemikiran & Cerita: AI

> Berbagi pengalaman, pembelajaran, dan petualangan saya di dunia teknologi dan lainnya.

URL: https://www.matthewswong.com/id/blog/category/ai/

- [Harness Engineering AI Agent: Bukti Riset dan Praktik Nyata](https://www.matthewswong.com/id/blog/ai-agent-harness-engineering-research/)
- [Cara Menulis File CLAUDE.md yang Efektif untuk Claude Code](https://www.matthewswong.com/id/blog/cara-menulis-claude-md-yang-efektif/)
- [Backstage MCP Actions Backend: Secure AI Access to Your Portal](https://www.matthewswong.com/id/blog/backstage-mcp-actions-backend-ai-agents/)
- [Stack Overflow Developer Survey 2026: Trust ke AI Agent](https://www.matthewswong.com/id/blog/stack-overflow-developer-survey-2026-ai-agent-trust/)
- [Studi Produktivitas Developer AI METR 2026: Apa Isinya](https://www.matthewswong.com/id/blog/metr-developer-productivity-2026-uplift-update/)
- [Perplexity SPACE: Sandbox Agent Firecracker, Warm Pool, E2B](https://www.matthewswong.com/id/blog/perplexity-space-firecracker-agent-sandbox/)
- [Docker Sandboxes sbx: Claude Code di MicroVM Cloud](https://www.matthewswong.com/id/blog/docker-sandboxes-cloud-sandboxes-sbx-guide/)
- [Tutorial kagent Kubernetes: AI Agent di Cluster Laptop](https://www.matthewswong.com/id/blog/kagent-ai-agents-kubernetes-tutorial/)
- [Kubernetes Agent Sandbox CRD: AI Agent Stateful Terisolasi](https://www.matthewswong.com/id/blog/kubernetes-agent-sandbox-crd-ai-agents/)
- [Kubernetes AI Conformance: Apa yang Dituntut KAR dari Platform](https://www.matthewswong.com/id/blog/kubernetes-ai-conformance-program-explained/)
- [Cloudflare Code Mode MCP: Dua Tool, 1.000 Token, dan Sandbox](https://www.matthewswong.com/id/blog/cloudflare-code-mode-mcp-two-tools-1000-tokens/)
- [Elastic nightshift AI SRE: Mulai dari Engine Investigation](https://www.matthewswong.com/id/blog/elastic-nightshift-ai-sre-four-engines/)
- [A2A vs MCP: Protokol AI Agent Mana yang Anda Butuhkan?](https://www.matthewswong.com/id/blog/a2a-vs-mcp-agent-protocols/)
- [Optimasi Biaya AI Agent: Anggarkan Per Task, Bukan Per Call](https://www.matthewswong.com/id/blog/ai-agent-cost-per-task-budgets/)
- [Evaluasi AI Agent: Cara Menguji Tool Call dan Trajectory](https://www.matthewswong.com/id/blog/ai-agent-evals-trajectory-testing/)
- [Panduan ChatGPT Plugins (Apps SDK): Bangun Plugin ERP di MCP](https://www.matthewswong.com/id/blog/chatgpt-plugins-apps-sdk-mcp-guide/)
- [Claude Managed Agents vs OpenAI Agents API: Perbandingan](https://www.matthewswong.com/id/blog/claude-managed-agents-vs-openai-agents-api/)
- [Tutorial Google ADK 2.0: Graph Workflow dan Multi-Agent](https://www.matthewswong.com/id/blog/google-adk-2-graph-workflows-tutorial/)
- [Panduan Middleware LangChain create_agent: PII, Limit, Fallback](https://www.matthewswong.com/id/blog/langchain-create-agent-middleware-guide/)
- [LangGraph vs CrewAI 2026: Mana untuk Agent Produksi?](https://www.matthewswong.com/id/blog/langgraph-vs-crewai-production-agents/)
- [Tutorial Framework Agent TypeScript Mastra: Workflow dan Evals](https://www.matthewswong.com/id/blog/mastra-typescript-agent-framework-tutorial/)
- [Microsoft Agent Framework 1.0: Panduan Migrasi Semantic Kernel](https://www.matthewswong.com/id/blog/microsoft-agent-framework-semantic-kernel-migration/)
- [Agent Builder OpenAI Ditutup: Migrasi Sebelum 30 Nov 2026](https://www.matthewswong.com/id/blog/openai-agent-builder-shutdown-migration/)
- [Tutorial OpenAI Agents API: Hosted Agent dan Sandbox](https://www.matthewswong.com/id/blog/openai-agents-api-hosted-sandbox-tutorial/)
- [OpenAI Agents API vs Agents SDK vs Responses API: Pilih Mana?](https://www.matthewswong.com/id/blog/openai-agents-api-vs-agents-sdk/)
- [Guardrails OpenAI Agents SDK: Cek Input, Output dan Tool](https://www.matthewswong.com/id/blog/openai-agents-sdk-guardrails-tripwires/)
- [Handoff vs Agents as Tools di OpenAI Agents SDK Dijelaskan](https://www.matthewswong.com/id/blog/openai-agents-sdk-handoffs-vs-agents-as-tools/)
- [Sandbox Agents OpenAI Agents SDK: Manifest, Docker, dan E2B](https://www.matthewswong.com/id/blog/openai-agents-sdk-sandbox-agents-manifest/)
- [OpenAI Agents SDK Sessions: Memory SQLite, Redis dan Postgres](https://www.matthewswong.com/id/blog/openai-agents-sdk-sessions-redis-sqlalchemy/)
- [Tutorial OpenAI Agents SDK TypeScript: Bangun Agent dengan Tool](https://www.matthewswong.com/id/blog/openai-agents-sdk-typescript-tutorial/)
- [OpenAI Agents SDK vs Google ADK vs Claude Agent SDK (2026)](https://www.matthewswong.com/id/blog/openai-agents-sdk-vs-google-adk-vs-claude-agent-sdk/)
- [Tutorial OpenAI ChatKit: Chat Agent Self-Hosted di React](https://www.matthewswong.com/id/blog/openai-chatkit-self-hosted-chat-ui/)
- [OpenAI DevDay 2026 untuk Developer: Sol, Ultrafast, Agents API](https://www.matthewswong.com/id/blog/openai-devday-2026-developer-announcements/)
- [OpenAI Dots Agents: Cara Kerja Agent ChatGPT yang Selalu Aktif](https://www.matthewswong.com/id/blog/openai-dots-always-on-chatgpt-agents/)
- [OpenAI Evals Dihentikan: Migrasikan Eval Anda ke Promptfoo](https://www.matthewswong.com/id/blog/openai-evals-shutdown-promptfoo-migration/)
- [OpenAI Realtime Voice Agent lewat SIP dengan gpt-realtime-2.1](https://www.matthewswong.com/id/blog/openai-realtime-voice-agent-sip-phone/)
- [OpenAI Responses API Remote MCP: Approval, Filter, Tunnel](https://www.matthewswong.com/id/blog/openai-responses-remote-mcp-approvals/)
- [OpenAI Skills API dan Hosted Shell: SKILL.md di Container](https://www.matthewswong.com/id/blog/openai-skills-api-hosted-shell/)
- [Playwright MCP vs Browser Use vs Stagehand: Browser Agent](https://www.matthewswong.com/id/blog/playwright-mcp-vs-browser-use-vs-stagehand/)
- [WhatsApp AI Agent Indonesia: Bot Pesanan UMKM dengan Agents SDK](https://www.matthewswong.com/id/blog/whatsapp-ai-agent-umkm-indonesia/)
- [Biaya Token LLM Bahasa Indonesia: 1.23x, Bukan 2-3x](https://www.matthewswong.com/id/blog/bahasa-indonesia-llm-token-cost/)
- [RAG Bahasa Indonesia: Full-Text Search Postgres Sering Meleset](https://www.matthewswong.com/id/blog/rag-bahasa-indonesia-search-retrieval/)
- [Speech-to-Text Bahasa Indonesia di Production: WER Whisper Nyata](https://www.matthewswong.com/id/blog/speech-to-text-bahasa-indonesia-production/)
- [Cara ChatGPT Astra Membangun Game Portofolio 3D Saya, Langkah demi Langkah](https://www.matthewswong.com/id/blog/chatgpt-astra-3d-portfolio-game-archipelago/)
- [A2A Protocol: Cara AI Agent Menemukan dan Delegasi ke Peer](https://www.matthewswong.com/id/blog/a2a-protocol-agent-interoperability/)
- [Adopsi AI Indonesia: Apakah Kita Sudah di Jalur yang Benar?](https://www.matthewswong.com/id/blog/indonesia-ai-adoption-right-path/)
- [Blokir Grok di Indonesia: Moderasi Gambar AI untuk Builder](https://www.matthewswong.com/id/blog/indonesia-grok-block-ai-moderation-lesson/)
- [Onboarding AI Coding Agent ke Codebase yang Asing](https://www.matthewswong.com/id/blog/ai-agent-onboarding-unfamiliar-codebase/)
- [Prompt Regression Testing di CI: Nilai Rate, Bukan Satu Jawaban](https://www.matthewswong.com/id/blog/prompt-regression-testing-ci-evals/)
- [Sahabat AI: Panduan Developer LLM Indonesia dan Biaya Token](https://www.matthewswong.com/id/blog/sahabat-ai-indonesian-llm-developer-guide/)
- [Hooks Deterministik vs Agent Judgement di Claude Code](https://www.matthewswong.com/id/blog/deterministic-hooks-vs-agent-judgement/)
- [AI Code Review sebagai Merge Gate CI: Apa yang Boleh Block](https://www.matthewswong.com/id/blog/ai-code-review-merge-gate-ci/)
- [Context Engineering untuk AI Coding Agent di Monorepo](https://www.matthewswong.com/id/blog/context-engineering-monorepo-ai-agents/)
- [Budget Token LLM untuk Tim Engineering Kecil: Ke Mana Larinya](https://www.matthewswong.com/id/blog/llm-token-budget-engineering-team/)
- [Go dan TypeScript: Strict Type untuk Coding dengan AI](https://www.matthewswong.com/id/blog/go-typescript-ai-assisted-coding/)
- [Advisor Tool Claude Code: Model Kedua di Momen Penting](https://www.matthewswong.com/id/blog/claude-code-advisor-tool-second-opinion/)
- [Auto Memory Claude Code: Apa yang Diingat Claude dan Kenapa](https://www.matthewswong.com/id/blog/claude-code-auto-memory-recall/)
- [Fast Mode dan Effort Level Claude Code: Cepat lawan Murah](https://www.matthewswong.com/id/blog/claude-code-fast-mode-effort-levels/)
- [Perintah goal Claude Code: Berputar Sampai Syarat Terpenuhi](https://www.matthewswong.com/id/blog/claude-code-goal-command-autonomous-loop/)
- [Dependensi Plugin Claude Code dan Batasan Versinya](https://www.matthewswong.com/id/blog/claude-code-plugin-dependencies-versioning/)
- [Plugin Marketplace Claude Code: Menulis marketplace.json](https://www.matthewswong.com/id/blog/claude-code-plugin-marketplace-authoring/)
- [Background Agent Claude Code: Model Supervisor-nya](https://www.matthewswong.com/id/blog/claude-code-agent-view-background-sessions/)
- [Channels Claude Code: Dorong Event ke Sesi yang Hidup](https://www.matthewswong.com/id/blog/claude-code-channels-external-events/)
- [Ultrareview Claude Code: Review Multi-Agent Terverifikasi](https://www.matthewswong.com/id/blog/claude-code-code-review-ultrareview/)
- [Claude Code di Chrome: Otomasi Browser dari Terminal](https://www.matthewswong.com/id/blog/claude-code-chrome-browser-automation/)
- [Context Window 1M Claude Code: Kapan Membantu, Kapan Merugikan](https://www.matthewswong.com/id/blog/claude-code-1m-context-window-strategy/)
- [Cross-Session Messaging Claude Code: Sesi yang Saling Bicara](https://www.matthewswong.com/id/blog/claude-code-cross-session-messaging/)
- [Konfigurasi Model Claude Code: Alias, Fallback, Effort](https://www.matthewswong.com/id/blog/claude-code-model-config-fallback-picker/)
- [Prompt Caching Claude Code: Setel TTL, Tekan Tagihan](https://www.matthewswong.com/id/blog/claude-code-prompt-caching-ttl-tuning/)
- [Tool Search Claude Code: Tunda Tool MCP, Hemat Konteks](https://www.matthewswong.com/id/blog/claude-code-tool-search-deferred-tools/)
- [Claude Code Agent Teams: Workflow Multi-Agent Paralel](https://www.matthewswong.com/id/blog/claude-code-agent-teams-guide/)
- [Claude Code Dynamic Workflows: Orkestrasi 1.000 Agent](https://www.matthewswong.com/id/blog/claude-code-dynamic-workflows-orchestration/)
- [Claude Code Remote Control: Kendalikan Sesi Lokal Jarak Jauh](https://www.matthewswong.com/id/blog/claude-code-remote-control-mobile-sessions/)
- [Data Pemakaian Claude Code: Temuan dari 400.000 Sesi](https://www.matthewswong.com/id/blog/claude-code-usage-statistics-research-2026/)
- [Rantai Fallback Model Gratis untuk Agen Otonom](https://www.matthewswong.com/id/blog/hermes-agent-free-tier-model-fallback-chain/)
- [Desain Prompt Agen Otonom: Commit Kecil yang Mendarat](https://www.matthewswong.com/id/blog/hermes-agent-prompt-design-tiny-commits/)
- [Memberi Memori pada Agen Coding Otonom: KNOWLEDGE + RAG](https://www.matthewswong.com/id/blog/autonomous-agent-knowledge-rag-loop/)
- [Setup MCP Server Claude Code: Scope, Auth, Budget Tool](https://www.matthewswong.com/id/blog/claude-code-mcp-server-setup-governance/)
- [Context Window Claude Code: Compaction dan Kendali Token](https://www.matthewswong.com/id/blog/claude-code-context-window-compaction/)
- [Claude Code di Monorepo: CLAUDE.md Bertingkat dan Skill](https://www.matthewswong.com/id/blog/claude-code-monorepo-workflow/)
- [Pipeline Konten Claude Code untuk Blog Next.js Dwibahasa](https://www.matthewswong.com/id/blog/claude-code-bilingual-blog-content-pipeline/)
- [Hooks Claude Code: Guardrail Agen yang Deterministik](https://www.matthewswong.com/id/blog/claude-code-hooks-guardrails-guide/)
- [Slash Command Claude Code: Otomatiskan Prompt Anda](https://www.matthewswong.com/id/blog/claude-code-slash-commands-guide/)
- [Skills Claude Code: Tulis Prosedur Sesuai Permintaan](https://www.matthewswong.com/id/blog/claude-code-skills-authoring-guide/)
- [Plan Mode Claude Code: Kurangi PR yang Rusak](https://www.matthewswong.com/id/blog/claude-code-plan-mode-guide/)
- [Cara Menulis CLAUDE.md yang Meningkatkan Output AI](https://www.matthewswong.com/id/blog/writing-claude-md-memory-guide/)
- [Checkpointing Claude Code: Mundurkan Perubahan Berisiko](https://www.matthewswong.com/id/blog/claude-code-checkpointing-rewind-guide/)
- [Optimasi Biaya Claude Code: Pangkas Konsumsi Token](https://www.matthewswong.com/id/blog/claude-code-cost-optimization-guide/)
- [Output Styles Claude Code: System Prompt Kustom](https://www.matthewswong.com/id/blog/claude-code-output-styles-guide/)
- [Pola Writer-Reviewer untuk Coding Agent AI](https://www.matthewswong.com/id/blog/ai-writer-reviewer-agent-pattern/)
- [Cara Memperbaiki Loop Handoff Tak Terbatas Agen AI](https://www.matthewswong.com/id/blog/ai-agent-handoff-loops-fix/)
- [Coding Agent AI Terminal Terbaik di 2026](https://www.matthewswong.com/id/blog/best-terminal-ai-coding-agents-2026/)
- [Claude Code: Skills vs Hooks vs Subagents vs MCP](https://www.matthewswong.com/id/blog/claude-code-skills-vs-hooks-vs-subagents/)
- [Claude Code vs Cursor vs Copilot: Adu 2026](https://www.matthewswong.com/id/blog/claude-code-vs-cursor-vs-copilot/)
- [Fan-Out vs Pipeline: Pola Agen AI](https://www.matthewswong.com/id/blog/fan-out-vs-pipeline-agent-pattern/)
- [Subagent Claude Code: Panduan Setup dan Alur Kerja](https://www.matthewswong.com/id/blog/claude-code-subagents-setup-guide/)
- [Plugin Claude Code Terbaik 2026: 62 Alat Alur Kerja Agentik](https://www.matthewswong.com/id/blog/claude-code-plugins-marketplace-guide/)
- [AI Product Description Generator SaaS for Sellers](https://www.matthewswong.com/id/blog/ai-product-description-saas-indonesia/)
- [Agentic Coding Workflow: Best Practices for AI Agents](https://www.matthewswong.com/id/blog/agentic-coding-workflow-best-practices/)
- [AI Agent vs LLM: What Is the Difference?](https://www.matthewswong.com/id/blog/ai-agent-vs-llm-model/)
- [Claude Code vs Codex: Agentic Coding Compared](https://www.matthewswong.com/id/blog/claude-code-vs-codex-agentic-coding/)
- [DeepSeek V4 Flash: Building a Spider-Man Cafe Tracker](https://www.matthewswong.com/id/blog/deepseek-v4-flash-cafe-tracker/)
- [How Claude Code Works: Agentic Coding Explained](https://www.matthewswong.com/id/blog/how-claude-code-works-agentic-coding/)
- [Human-in-the-Loop Design for AI Agents](https://www.matthewswong.com/id/blog/human-in-the-loop-ai-agents/)
- [MCP OAuth: Why Authentication for AI Agents Is Hard](https://www.matthewswong.com/id/blog/mcp-oauth-authentication-hard/)
- [MCP vs Function Calling: How AI Agents Use Tools](https://www.matthewswong.com/id/blog/mcp-vs-function-calling/)
- [Multi-Agent Orchestration: Subagents Explained](https://www.matthewswong.com/id/blog/multi-agent-orchestration-subagents-workflow/)
- [Spec-Driven Development With AI Coding Agents](https://www.matthewswong.com/id/blog/spec-driven-development-ai-agents/)
- [Cloudflare Wallets: Programmable Payments for AI Agents](https://www.matthewswong.com/id/blog/cloudflare-wallets-agentic-payments/)
- [Context Engineering untuk Agen LLM: Write, Select, Compress](https://www.matthewswong.com/id/blog/context-engineering-llm-agents/)
- [Jalankan LLM Lokal dengan Ollama: Model, API, Modelfile](https://www.matthewswong.com/id/blog/ollama-run-local-llms/)
- [Panduan Vercel AI SDK 5: Chat Type-Safe, Tool & Agen](https://www.matthewswong.com/id/blog/vercel-ai-sdk-5-guide/)
- [Mengenal llms.txt: Standar AI untuk Konten Situs Web](https://www.matthewswong.com/id/blog/llms-txt-ai-website-standard/)
- [Output Terstruktur LLM: JSON Schema yang Benar-Benar Tervalidasi](https://www.matthewswong.com/id/blog/llm-structured-output-json-schema/)
- [LLM Evals di CI: Menguji Prompt Seperti Kode](https://www.matthewswong.com/id/blog/llm-evaluation-testing-ci-pipeline/)
- [Membangun Agen dengan Claude Agent SDK](https://www.matthewswong.com/id/blog/claude-agent-sdk-building-agents/)
- [Semantic Caching untuk LLM: Hemat Biaya Pertanyaan Berulang](https://www.matthewswong.com/id/blog/semantic-caching-llm-cost-reduction/)
- [Ekstraksi Data Dokumen dengan Multimodal LLM](https://www.matthewswong.com/id/blog/multimodal-llm-document-data-extraction/)
- [Membangun Chatbot Streaming dengan Groq dan Next.js](https://www.matthewswong.com/id/blog/groq-nextjs-streaming-chatbot/)
- [Claude Fable 5: Yang Perlu Developer Ketahui](https://www.matthewswong.com/id/blog/claude-fable-5-developers-guide/)
- [LLM Prompt Caching: Optimasi Biaya Termurah yang Belum Anda Pakai](https://www.matthewswong.com/id/blog/llm-prompt-caching-cost-optimization/)
- [AI Agent di Produksi: Pelajaran Setahun Workflow Agentic](https://www.matthewswong.com/id/blog/ai-agents-production-lessons/)
- [MCP Dijelaskan: Apa yang Sebenarnya Diselesaikan Model Context Protocol](https://www.matthewswong.com/id/blog/mcp-model-context-protocol-explained/)
- [Cara Saya Menggabungkan Opus 4.7 dan Sonnet 4.6 di Claude Code](https://www.matthewswong.com/id/blog/claude-opus-4-7-whats-new/)
- [Hermes AI Agent oleh Nous Research: Arsitektur, Loop Belajar Mandiri, dan Cara Memulai](https://www.matthewswong.com/id/blog/hermes-ai-agent-nousresearch-guide/)
- [Otorisasi MCP dengan Cerbos: Policy-as-Code untuk AI Agent](https://www.matthewswong.com/id/blog/mcp-auth-cerbos-authorization/)
- [Cara Saya Memangkas Tagihan LLM API 60% Tanpa Mengorbankan Kualitas Model](https://www.matthewswong.com/id/blog/llm-api-cost-optimization/)
- [Otomatisasi Alur Kerja AI Agent di 2025: Yang Benar-Benar Berhasil](https://www.matthewswong.com/id/blog/ai-agent-workflow-automation-2025/)
- [Pelajaran Produksi Pipeline RAG: Yang Tidak Diceritakan Siapa Pun](https://www.matthewswong.com/id/blog/rag-pipeline-production-lessons/)
- [Claude API Tool Use di Produksi: Panduan Dunia Nyata](https://www.matthewswong.com/id/blog/claude-api-tool-use-production/)
- [OpenAI vs Anthropic vs Google Gemini: Perbandingan Jujur Developer (2025)](https://www.matthewswong.com/id/blog/openai-vs-anthropic-vs-gemini-developers/)
- [LangChain vs Membangun Orkestrasi LLM Sendiri: Kapan Menggunakan Masing-Masing](https://www.matthewswong.com/id/blog/langchain-vs-custom-llm/)
- [Code Review Bertenaga AI dengan GitHub Actions: Setup Saya](https://www.matthewswong.com/id/blog/ai-code-review-github-actions/)
- [pgvector vs Pinecone vs Qdrant: Memilih Database Vektor di 2025](https://www.matthewswong.com/id/blog/pgvector-vs-pinecone-comparison/)
- [Rekayasa Prompt untuk Produksi: Di Luar Dasar-Dasar](https://www.matthewswong.com/id/blog/prompt-engineering-production/)
- [Kontrol Kualitas Pembuatan Konten AI: Yang Saya Pelajari dengan Cara Sulit](https://www.matthewswong.com/id/blog/ai-content-generation-quality/)
- [Pertahanan Prompt Injection: Cara Saya Mengamankan Sistem AI dari Manipulasi](https://www.matthewswong.com/id/blog/ai-prompt-injection-defense/)
- [Membangun MCP Server dari Nol: Panduan Praktis untuk Developer](https://www.matthewswong.com/id/blog/mcp-server-development-guide/)
- [Pola Persistensi Memori Agen AI: Dari Volatile hingga Jangka Panjang](https://www.matthewswong.com/id/blog/ai-agent-memory-persistence-patterns/)
- [Menjalankan LLM Lokal dengan Ollama di Produksi: Apa yang Saya Pelajari](https://www.matthewswong.com/id/blog/local-llm-ollama-production/)
- [Observabilitas AI: Cara Log, Trace, dan Monitor Aplikasi LLM](https://www.matthewswong.com/id/blog/ai-observability-llm-logging/)
- [Pola Function Calling: OpenAI vs Anthropic Tool Use di Produksi](https://www.matthewswong.com/id/blog/function-calling-openai-anthropic/)

---

- HTML version: https://www.matthewswong.com/id/blog/category/ai/
- English: https://www.matthewswong.com/en/blog/category/ai/
- Site index for agents: https://www.matthewswong.com/llms.txt
- Sitemap: https://www.matthewswong.com/sitemap.xml
- API description (OpenAPI): https://www.matthewswong.com/openapi.json