Security in the age of AI 🔗
AI agents are finding old vulnerabilities and making complex attacks easier. A short history of automated hacking, what comes next, and how to prepare [...]
Flavio Copes
Posts
Give your coding agent a state machine 🔗
Build a small JavaScript state machine for a coding agent, with explicit transitions, a retry limit, a useful event log, and tests. [...]
PSA: update PHP today 🔗
PHP security fixes are available. Update now, check the fixed versions, and use these temporary mitigations only while waiting for patched packages. [...]
Jev vs structured outputs: when to use each 🔗
Compare Jev, JSON mode, and schema-constrained LLM outputs. Learn when a typed answer is enough and when decision probabilities justify another model. [...]
Vite toolchain: Cloudflare, vinext, and shared web tooling 🔗
Vite is becoming shared infrastructure for frameworks, runtimes, and developer tools. What vinext, Cloudflare's new tooling, and DevTools show. [...]
SPA vs server-side rendering: which should you choose? 🔗
Choose between a single-page app, server-rendered pages, and interactive islands by looking at navigation, shared state, loading, and maintenance. [...]
A deep dive into tsrs 🔗
Understand tsrs, the Rust port of TypeScript. Try the compiler, compare benchmarks, explore its architecture, and build a browser-based type checker. [...]
How to update all global npm packages 🔗
Update all global npm packages with npm update -g, check outdated tools, update one package, and understand which Node installation you are changing. [...]
Cloudflare Pages vs Workers: which should you choose? 🔗
Compare Cloudflare Pages and Workers for static sites and full-stack apps, with request diagrams, pricing examples, and advice on when to migrate. [...]
AI-assisted code migration: how to switch frameworks safely 🔗
Coding agents can handle repetitive migrations, but switching costs also include design drift, production behavior, deployment, and future maintenance [...]
How AI agents make existing software more useful 🔗
Agents can help us combine existing tools and improve the details of our apps. That makes dependable behavior and clear interfaces worth paying attent [...]
How I would evaluate Muse Wonder in Cursor 🔗
Muse Wonder is available in Cursor. Here's how to compare its coding results, reasoning effort, and token costs against the model you already use. [...]
How to compare OpenCode, Claude Code, and Codex CLI 🔗
Compare OpenCode, Claude Code, and Codex CLI with one reproducible bug fix, independent tests, setup checks, and an honest look at cost and interventi [...]
How to speed up tests for AI coding agents 🔗
How to shorten an agent's feedback loop with focused tests, realistic boundaries, useful failures, and broader checks before shipping. [...]
15 apps you can vibe code instead of subscribing 🔗
Build your own Slack, Notion, Calendly, and more. These 15 guides include scoped AI prompts, realistic build estimates, and smaller features to start [...]
AI coding agent context: how to document project decisions 🔗
A practical way to preserve project decisions, rejected alternatives, and unfinished questions so you and your coding agents can pick up the work late [...]
A practical review workflow for T3 Code threads 🔗
Use T3 Code's stable release with a clear workflow for isolated agent changes, reviewing the complete branch, merging, and cleaning up finished work. [...]
How to use OpenCode with a local model in LM Studio 🔗
Connect OpenCode to LM Studio, test local tool calling, fix model discovery problems, and check what you need for a coding workflow that works offline [...]
Parallel coding agents: how to coordinate and review changes 🔗
How I think about parallel coding agents, shared code, review queues, and the checks that tell us whether their changes work together. [...]
Codex vs Claude Code vs Cursor: how to choose 🔗
How I would compare coding agents by finished work, usage limits, review effort, and switching costs, while keeping the project portable. [...]
Changing package managers doesn't remove your npm registry dependency 🔗
npm is both a command and a package registry. Learn what switching to pnpm or Bun changes, and how caches, lockfiles, and mirrors affect failed instal [...]
OpenCode Code Mode explained: JavaScript instead of repeated tool calls 🔗
Learn how OpenCode Code Mode combines tool calls in JavaScript, when batching helps, and where permissions, errors, and side effects still need care. [...]
MCP server design: tools, authentication, and rate limits 🔗
How to design a useful MCP interface for your product, with clear tools, account permissions, usage limits, predictable results, and safe writes. [...]
Selling SaaS to developers: how to find paying customers 🔗
Developers can build alternatives to your product. That doesn't make them bad customers. Look at urgency, budgets, and the work buying lets them avoid [...]
How to build agents that survive crashes with Pi Durable 🔗
Build a durable TypeScript agent with Pi Durable and SQLite. Test crash recovery, avoid duplicate requests, and understand which tools can safely run [...]
A deep dive into OpenSEO 🔗
Explore OpenSEO's architecture: SEO data APIs, Cloudflare Workers, its audit crawler, MCP tools, AI agents, storage, and self-hosting trade-offs. [...]
How to write an OpenCode plugin with the v2 API 🔗
Build your first OpenCode v2 plugin in TypeScript, register a custom tool, validate its inputs, test the logic, and understand how v1 plugins change. [...]
Build vs buy software: when AI can replace a SaaS 🔗
AI changes the build-versus-buy calculation, but a working prototype is only part of it. How to compare ownership, maintenance, risk, and recurring co [...]
How to compete with AI-generated SaaS clones 🔗
Cheap software clones change what a small SaaS business can charge for. A practical look at differentiation, maintenance, distribution, and earning tr [...]
Meta released Muse Spark Open: what you can run, and what it takes 🔗
Meta released Muse Spark Open with downloadable weights, a million-token context, and support for coding agents. Here are the hardware and license lim [...]