Daily Vibe - 10-22-26 · Vibe Coders

12d • News

Daily Vibe - 10-22-26

Good news: shipping fast with AI is still the move. Bad news: "it runs" and "it's safe" are two different checklists, and most of us have only been doing the first one.

The lead story: a production checklist for apps you vibe-coded

SitePoint published a developer checklist this week for taking AI-generated apps from prototype to production, and the reason it exists is the stat buried in the middle: Veracode's 2025 GenAI Code Security Report found that 45% of AI-generated code samples failed security tests, tripping OWASP Top 10 issues. That's Veracode's own benchmark on code samples, not full apps or real breaches, so take it as a signal, not a verdict. Separately, a WIRED investigation from May found exposed data in apps built with popular AI coding tools in the wild.

None of this means vibe coding is broken. It means the part where you turn a prompt into a login page and a dashboard was never the hard part. The checklist's actual advice is unglamorous: write down who can see what before you generate more features, ship one feature at a time instead of one giant prompt, and read the auth code specifically because a hidden button on the frontend does nothing if someone can just call your API directly. The example in the article is a classic: an app that lets a user view invoices needs to check that this user is allowed to see that invoice, server-side, every time. Not "logged in," but "logged in and allowed."

If you've shipped something with Cursor or Claude Code and haven't gone back to check that every protected endpoint actually enforces permissions per record, that's probably this week's homework.

[YOUR TAKE: have you run a security pass on something you vibe-coded, or found out the hard way?]

Agent Skills just became official MCP plumbing

Bloomberg helped standardize a spec called Skills over MCP, now an official MCP extension (SEP-2640), announced at AGNTCon + MCPCon Europe in Amsterdam. The short version: MCP servers can now serve both tools and the instructions for using them, as skills, from one place. An agent loads just the skill's name and description first, then pulls the full instructions only when it actually picks that skill. Update the skill on the server, every connected agent gets the new version automatically, no re-syncing files to every user. Bloomberg says it hit this problem building ASKB, its Terminal AI tool, once skills needed to scale past one team.

If you're building anything with MCP and keep hitting the "do I cram everything into context or limit what my agent can do" wall, this is the fix that's about to become standard. Worth a look before you build your own version of this from scratch.

OpenClaw's creator spent $10 million on tokens last month

Peter Steinberger, creator of OpenClaw and now at OpenAI, told Cloudflare Connect 2026 he spent roughly $10 million on AI tokens in September, covered by investors. When the room reacted, his response was blunt: "not all of that is useful." He also said AI infrastructure hasn't caught up, questioning why workloads still run locally instead of in the cloud, and floated a fix for giving agents too much permission: have one agent negotiate permissions with another before letting it act.

The number is absurd on purpose, this guy has what Cloudflare's CTO called a "virtually unlimited token budget." But the admission underneath it is the useful part: even with infinite money, a huge chunk of agent spend is waste. If you're running agents on a real budget, that's worth remembering before you let one loop unsupervised overnight.

[YOUR TAKE: how do you keep agent token spend from quietly spiraling?]

What this means for you

Before you call anything "done," check that sensitive actions (auth, payments, file access, anyone else's data) are enforced on the server, not just hidden in the UI.
Write a short spec before you generate the next feature: who can do what, what happens on failure, what counts as success. Feed that to your assistant instead of a vague prompt.
Keep an eye on Skills over MCP if you're building with MCP servers, it's the standardized answer to "how do I share reusable agent instructions without blowing up context."
If you're running agents on a loop or in the background, check in on actual token spend before it checks in on you.

What's the scariest thing a security pass has turned up in something you built fast? Drop it in the thread.

Sources