AI coding guides & resources
Engineering decisions, tutorials, and video guides from the OTF team.
iPhone Duo app readiness: layout testing and App Store screenshots
Prepare for iPhone Duo with an app-layout audit, resize tests, and a checklist for App Store screenshots ahead of the 2027 requirement.
8 min read
Lovable usage limits: choose an alert or block with live-app impact in view
Learn how Lovable’s credit-limit scope and type determine whether a rule alerts admins, blocks building, or pauses features in a live app.
8 min read
How to test whether coding agents discover the skills you need
Test whether coding agents find and use your skills from realistic prompts, traces, and a crowded catalog.
8 min read
OpenRouter says its Stripe announcement needs no integration change
OpenRouter said existing integrations would not change. Check your app’s request path, keys, data terms, and billing records before making assumptions.
8 min read
Anthropic’s November 12 Usage Policy update: checklist for AI app builders
A source-bounded checklist to review Claude app flows before Anthropic’s Usage Policy takes effect November 12, 2026.
6 min read
Claude Sonnet 5.5 cache reads cost half: measure your own savings
Anthropic cut Sonnet 5.5 cache-read pricing in half. Learn how to isolate the rate effect, log cache usage, and measure your own bill before changing prompts.
6 min read
Bolt.new alternatives: choose a path to ship a production app
Compare Bolt.new alternatives by code handoff, deployment control, mobile fit, and cost before you move a production app.
8 min read
A safe evaluation plan for Vercel’s free Glyph Cluster model
Glyph Cluster is free in stealth, but its anonymous provider, training warning and missing ZDR make a safe evaluation plan essential.
8 min read
Claude Haiku 5.5 production routing: measure cost, latency, and task quality
Choose high-volume tasks for Claude Haiku 5.5, account for token and price tiers, and validate quality, latency, and cost before rollout.
6 min read
Migrate three OpenAI API model IDs before the April 2027 shutdown
A practical inventory, evaluation, and staged rollout plan for replacing three OpenAI API model IDs before their April 2027 shutdown.
6 min read
Add file-based voice input and spoken replies to your OpenRouter app
Build file-based voice input and replies with Scribe v2 and Eleven v4 Turbo; measure both costs, latency, and transcript quality.
8 min read
Can you import a GitHub repository into Lovable? Here is the current answer
Lovable's GitHub sync creates a repository from a Lovable project. See why an existing GitHub repo cannot be imported and what to do instead.
7 min read
Vercel AI Gateway confidence fallbacks: measure before rollout
Measure trigger rate, decision quality, added latency, and second-stage cost before using confidence-based fallbacks in production.
8 min read
Mistral Large 4 hosted preview vs weights: choose by your constraints
Choose Mistral Large 4’s hosted preview only if data, latency, and service requirements fit; wait for the planned weights when self-run control is mandatory.
7 min read
Vercel Agent: shared registry credentials for private packages
Use NPM_TOKEN or NPM_RC as team-shared Development/Preview variables so Vercel Agent can install private packages without exposing credentials.
6 min read
OpenAI text watermarking does not give your app a provenance log
Keep app-owned records of who requested and changed generated text; watermarks cannot identify users or prove authorship.
7 min read
What you hold after a $9 site purchase, line by line
What a $9 landing template includes, what it leaves out, and how its price and license compare with Shadcnblocks and Framer.
7 min read
What runs before withSupabase in an MCP pipeline?
The 1.0 Supabase middleware runs OAuth discovery ahead of the auth gate. See what sits before withSupabase and what follows it.
6 min read
Which addresses go on cloudflared --allowed-mail?
cloudflared 2026.9.3 adds --allowed-mail to Quick Tunnels. Who to list, what to put in AGENTS.md, and when to move to Access.
7 min read
Does anything still expect a 40-character GitHub token?
GitHub App tokens went from 40 to about 520 characters. Check columns, proxies and log rules, and the 30 Nov header deadline.
6 min read
What a Claude Code mod can do with your permissions
A guide to Claude Code mod permissions, the events and API calls to inspect, and the org settings that decide which mods load.
8 min read
Can you swap a Jev decision call for Clef on Workers AI?
Cloudflare says Clef is Jev-API compatible. See what changes, what its tables report, and how to test a swap on your decisions.
7 min read
Use Laya for structured decisions in your app
Learn what Laya returns, where yes/no, routing, and scoring questions fit, how to try it on AI Gateway, and what to plan by Oct 31.
8 min read
Fitness app template for trainers: the $99 kit
Pre-buy fit check for a trainer’s member app: see what the $99 Fitness Kit covers and what you still need to build yourself.
7 min read
