Topic
Building with AI tools, no-code, and the new vibe-coding workflow.
Jun 4, 2026 · 10 min read
The seven stage-of-product signals that tell you your AI-built app has outgrown DIY. Regulated data, SLA commitments, real onboarding, multi-region, diligence, security reviews, audits.
Read more
May 31, 2026 · 10 min read
How to structure an agency retainer when you ship daily with Lovable or Bolt but need humans for deploy, audits, incidents, and monthly ops. Pricing, scope, and the clauses that matter.
May 28, 2026 · 11 min read
AI tools generate code that works today but is hard to modify tomorrow. The five categories of AI-specific tech debt we find most, and the refactor patterns that pay them down without a full rewrite.
May 25, 2026 · 10 min read
The package a founder should send to an agency that is picking up a Lovable project. Repo access, Supabase credentials, env vars, design tokens, roadmap, known issues, and the secrets rotation nobody...
May 21, 2026 · 11 min read
AI coding tools get a founder to 70 percent in an afternoon. The last 30 percent is where every production incident lives. This is the category, the skill stack, and why neither a full-stack...
May 18, 2026 · 10 min read
Plan mode generates a full plan before writing code. For non-technical founders, this is the most important Claude Code feature. How to use it, what to review, when to stop it and restart.
May 15, 2026 · 9 min read
Three agentic coding tools tested on the same four projects. Output quality, cost, speed, human intervention needed. Decision framework by task type.
May 11, 2026 · 9 min read
The skills directory structure (what you keep there, what you don't), the hooks config, and the three workflows we have automated this way across client projects.
May 8, 2026 · 10 min read
Every Monday: dependency update, security patch, test run, log review, cost review. The Claude Code automation we run to keep client projects from rotting between feature work.
May 4, 2026 · 10 min read
Start to finish: prompt, generate, local-run, commit, Vercel connect, env vars, DNS, live. The shortest path we know. For solo founders who want a URL by end of day.
May 1, 2026 · 10 min read
Breaking a large task into sub-agents that each have narrow context. The pattern we use for 8-hour refactors, data migrations, and cross-repo changes. Includes actual prompt templates.
Apr 27, 2026 · 11 min read
The project-level instructions file that transforms Claude Code from helpful to operational. The WitsCode CLAUDE.md template, what to include, what to exclude, and the commit-message rule that saves...
Apr 24, 2026 · 11 min read
Proration on upgrade, cancel at period end versus immediate, trial expiry, failed payments, and the reactivation door that swings one way. The Stripe subscription paths your AI builder never wrote.
Apr 23, 2026 · 10 min read
AI-generated schemas rarely include indexes. Walk through the EXPLAIN ANALYZE workflow, the four index types we add most, and the migration pattern to add them without downtime.
Apr 21, 2026 · 10 min read
22 checks before you flip Stripe to live mode. Webhook endpoints, test cards, tax, legal pages, invoice branding, refund policy, dispute handling.
Apr 17, 2026 · 11 min read
If your app moves money on behalf of other users, Connect is mandatory and Checkout will not save you. A walk through the account-type decision, the AccountLink onboarding flow, application fees, and...
Apr 15, 2026 · 10 min read
The reliable-webhook pattern for Stripe: signature verification with raw bodies, event.id idempotency, async 200-OK-then-process, and dead-letter queues. Why passing test mode is not the same as...
Apr 11, 2026 · 11 min read
Stripe is easy to wire up in a demo and hard to get right in production. The nine edge cases AI tools miss, from raw-body webhooks to dispute alerts, and how to fix each one.
Apr 7, 2026 · 10 min read
Magic links are the default on Lovable. They are not always right. Here is when password auth wins, when OTP makes sense, and the hybrid pattern we use on high-trust B2B apps.
Apr 4, 2026 · 10 min read
AI-built Next.js apps almost always protect routes in React, which is not protection at all. Here is the server-side middleware layer with Supabase SSR that actually stops bad actors, plus the...
Apr 1, 2026 · 9 min read
Session Handling AI Tools Get Consistently Wrong When an AI coding tool builds you an authenticated app, the login screen almost always works. A user types an email, receives a magic link or enters a...
Mar 28, 2026 · 11 min read
Lovable ships auth but not RBAC. The role schema, custom access token hook, RLS policies and server-side UI checks that actually hold up in production.
Mar 25, 2026 · 10 min read
Three realistic auth options for founders who have outgrown the default. A decision framework covering cost at 10K MAU, SSO readiness, Organizations, and the real price of DIY.
Mar 22, 2026 · 11 min read
Signals you have outgrown Supabase and the honest migration path to managed Postgres plus a real auth provider, including the zero-downtime playbook most guides get wrong.
