AI agents, Cloudflare systems, and product engineering
I’m Surya Setiyaputra, a software engineer in Sydney building AI agent systems, developer tools, and internet products.
Start here
A2A
Building with A2A: Agent-to-Agent Protocol in Practice
A practical primer on how A2A works, when to use it, and how agents discover and delegate tasks.
Memory
Persistent memory for AI agents
How I built a recovery-friendly memory system for long-running agents with layered backups and integrity checks.
Security
Agent SEO and prompt-injection risk
Why web-facing agents are vulnerable to hidden instructions and what practical defenses matter.
The agent's journey — roll the dice, climb ladders, slide down snakes.
Recent GitHub Activity
2026
- How Cron Prompts Evolved After Migrating from OpenClaw to Hermes ai-agentsopenclawhermescronautomationmigration
A practical comparison of how scheduled AI agent prompts changed after migrating from OpenClaw to Hermes: transport, state, runtime boundaries, and script-first automation.
16 min read - From OpenClaw to Hermes: How I Retired an Agent Home Without Losing the Archive AIAgentic AIOpenClawHermesmigrationDevOpsautomation
How to retire an OpenClaw home after moving to Hermes without losing archive access, breaking cron jobs, or leaving stale git hook paths behind.
11 min read - Compiling a Brag Doc: Engineering Growth Evidence in the Agentic IDE Era AIAgentic AIEngineeringCareer GrowthCursorKnowledge Management
Two months of running a Cursor-agent-maintained vault changed how I think about brag docs. The capture problem is mostly solved by the IDE; the missing piece is synthesis. Here's the layered system I built and what I'd tell you if you're starting.
8 min read - What Factual AU's Narrative Clusters Reveal About Budget 2026 Discourse Public PolicyData VisualizationMediaTaxOpen Source
The recurring narratives in Factual AU show how quickly scenario-dependent Budget 2026 claims get compressed into deterministic slogans.
8 min read - A Budget 2026 Matrix for Real Investor Decisions AIPublic PolicyData VisualizationOpen SourceSide Project
I built a public matrix that turns abstract Budget 2026 tax claims into concrete scenario-by-scenario decisions across investor archetypes and life stages.
6 min read - A2A Matters Most at the Boundaries Between Systems AIA2AInteroperabilityProtocolsMulti-Agent Systems
Why the real promise of A2A is cross-system and cross-organization interoperability, where agents need a shared contract without sharing internals.
7 min read - When the Keys Stop Working: Migrating an AI Agent System from Anthropic to OpenAI AIAgentic AIAutomationOpenClawDevOpsOpenAI
Anthropic blocked OAuth access for third-party tools. OpenAI still supports it. Here's how I migrated 48 cron jobs and a multi-agent system — and the future-proofing strategy that came out of it.
8 min read - OpenClaw Part 3: How My AI Agent's Memory Evolved Over Two Months AIAgentic AIAutomationOpenClawDevOps
From flat files to tiered memory architecture — how two months of running AI agents taught me that memory isn't a feature, it's infrastructure. Includes multi-agent sharing, self-healing systems, and automated DR.
12 min read - Launching indonesia-civic-stack: 40 MCP Tools for Indonesian Government Data AIMCPOpen SourcePythonCivic Tech
How I built an open-source Python SDK and MCP server that connects AI agents to 11 Indonesian government portals — and what I learned along the way.
5 min read - TaskFolio: Seeing Which Parts of Your Job AI Will Actually Affect AIJobsAutomationSide ProjectOpen Source
I built a tool that breaks down AI's impact on Australian jobs at the task level — not just 'your job is 40% automatable' but exactly which tasks, when, and how.
4 min read - Launching Agent Taxonomy: A Pokédex for AI Agents AIAgentic AIProduct HuntSide ProjectA2A
Today I'm launching Agent Taxonomy on Product Hunt — a playful species classifier that gives your AI agent a Latin binomial name, rarity tier, and a Pokémon-style portrait.
5 min read - Bypass Geo-Blocking with Cloudflare Workers: Reverse Proxy Guide CloudflareAIProxyInfrastructure
Step-by-step guide to bypass geo-blocking with a Cloudflare Worker reverse proxy. Useful for geo-restricted APIs, testing, and edge routing.
7 min read - How to Monetize AI Agent Skills — From Repo to Revenue AIBusinessSkillsMonetizationIndie Dev
Real numbers from selling agent skills: what works, what fails, and why paid ads kill margins. Multi-channel strategy for low-cost digital products.
7 min read - Building Price Intelligence from 118K Grocery Data Points PythonData EngineeringWeb ScrapingAnalyticsSide Projects
Turn grocery price data into a system that detects fake sales and shrinkflation. Built with Python, SQLite, and basic statistics in 118K data points.
6 min read - Building a Personal Finance Site for Indonesia — Launch Story Side ProjectsFinanceSEOAstroCloudflare
From zero to 89 bilingual articles: building a passive investing education site with a local retirement calculator and SEO-first architecture.
7 min read - Building a2alist.ai: A Protocol Directory for the Agentic Web AIA2Ax402CloudflareAgentsSide Projects
How I built a discovery platform for A2A and x402 protocol implementations — featuring automated security reviews, an x402-powered Discovery API, and the infrastructure that lets agents find other agents.
10 min read - Multi-Agent Content Pipelines: Practical Orchestration Patterns AIAgentic AIContentAutomationMulti-Agent
How I run a content pipeline with specialized AI agents — researcher, writer, editor, publisher — and the orchestration patterns that actually work at scale.
7 min read - Recursive Patterns for Scaling Agent Work AIAgentic AIResearchPatternsMulti-Agent
Three patterns from the Recursive Language Models paper that changed how I orchestrate multi-agent systems — and you don't need a research lab to use them.
7 min read - x402 as Anti-Spam: Micropayments as a Quality Gate x402AISpam PreventionProtocolsCrypto
Forget monetization — x402's most practical use case might be as a spam filter. How a $0.99 payment requirement eliminates low-quality submissions without blocking legitimate users.
7 min read - Building with A2A: Agent-to-Agent Protocol in Practice AIA2AAgentic AIProtocolsMCP
A practical architecture guide to Google's Agent-to-Agent (A2A) protocol, including agent cards, task flow, lifecycle management, and where A2A fits next to MCP and x402.
14 min read - Building Skills for Claude: A Practical Guide AIClaudeSkillsAgentic AIOpenClaw
Anthropic's official patterns for building Claude skills. Progressive disclosure, trigger design, multi-agent workflows from 16 production skill upgrades.
9 min read - Setting Up Tailscale for Your OpenClaw Instance tailscaleopenclawnetworkingsecuritytutorial
Securely expose your OpenClaw AI agent gateway with Tailscale. Zero-config mesh networking with no open ports or public IP required. Step-by-step setup.
10 min read - How to Make AI Agent Skills Portable and Reusable ai-agentsskillsportabilitypinanobotmigration
Build AI agent skills once, use everywhere. Learn the SKILL.md standard that makes capabilities portable across platforms with real migration examples.
10 min read - How to Migrate OpenClaw to a New Server (2026) — Step-by-Step Guide openclawai-agentsmigrationmemory-persistencedevops
Complete guide to migrate OpenClaw to a new server or platform. Transfer memory, cron jobs, secrets, and skills to IronClaw, Nanobot, Pi, or Goose without losing data.
16 min read - x402: Monetizing Content for AI Agents with HTTP 402 AIx402CryptoUSDCCloudflare WorkersMonetization
How to set up x402 payment protocol on a static site — letting AI agents pay USDC per API call while humans browse for free. A practical walkthrough using Cloudflare Workers and Base.
8 min read - How to Set Up ProtonVPN on Raspberry Pi as a VPN Gateway (2026 Guide) raspberry pivpnwireguardprotonvpnnetworkingprivacytutorial
Step-by-step guide to configure ProtonVPN on a Raspberry Pi using WireGuard. Route your smart TV, gaming console, and IoT devices through a secure VPN gateway for privacy.
7 min read - Agent SEO Is Social Engineering — How AI Reads Your Website AISecurityAgentic AISocial Engineering
Attackers are embedding hidden instructions in web pages to manipulate AI agents. Learn how agent-targeted SEO works and how to defend against it.
6 min read - OpenClaw Part 2: Persistent Memory and Disaster Recovery AIAgentic AIAutomationOpenClaw
How I built a memory system for AI agents that survives server failures — with tiered backups, automated restoration, and lessons from the agentic AI community.
6 min read - Experimenting with OpenClaw: Running a Multi-Agent AI System AIAgentic AIAutomationOpenClaw
Practical lessons from setting up and operating a multi-agent AI system — architecture decisions, cost management, and the surprising compound effects of small automations.
4 min read - Hello (Again), World! metabloggingai
From WordPress to AI-built static site: how I shipped a complete blog revamp with 98% AI-written content, prompted entirely through messaging apps.
2 min read
2025
- 2025 Portfolio Review: The Year of Consolidation and Compounding Personal FinanceInvesting
11% net worth growth through consolidation and discipline. FIRE target hits 43% as the portfolio compounds without constant tinkering. 2025 review.
5 min read