← The AI Hype Audit — all 217 verdicts

PARTLY “Head of Claude Code: I’m not prompting agents anymore, I’m building graphs and loops” — the loops are real; the quote and the buzzword are dressed up X · Aug 2026

The claimA viral post quoting the “Head of Claude Code” saying “I’m not prompting my agents anymore, I’m building graphs and loops so they can build the agents for me,” promoting a 10-minute talk and a step-by-step “graph engineering” guide, with the tagline “most people use AI at 5–10% of what it can do.”

There’s a real person and a real idea underneath the polish. Boris Cherny genuinely is the creator and head of Claude Code, and he really did say (June 2026, covered by The New Stack, Fortune and others) “I don’t prompt Claude anymore. I have loops running that prompt Claude … my job is to write loops.” That shift — from hand-typing prompts to running standing agent loops — is real and worth understanding. But the post embellishes: Cherny said LOOPS; “graphs” and “so they can build the agents for me” are spliced in from a separately-trending “graph engineering” meme he’s not on record saying. “Graph engineering” itself is a mid-2026 community label slapped onto multi-agent patterns Anthropic already published — real technique, manufactured branding. And “5–10% of what AI can do” is the same floating, unmeasurable hook this account reuses to funnel readers to its guides.

What holds up

  • Boris Cherny really is the creator/head of Claude Code (Lenny’s Podcast, Anthropic pages)
  • The loops quote is real — from a June 2026 interview, covered by The New Stack and Fortune
  • Standing agent loops and multi-agent orchestration are genuine, documented techniques (Anthropic’s own ‘Building Effective Agents’)

What doesn’t

  • The quote is doctored — Cherny said ‘loops,’ not ‘graphs and loops so they can build the agents for me’
  • ‘Graph engineering’ is a ~week-old community coinage, not a Cherny doctrine or an Anthropic term
  • “10-minute talk” is a clipped excerpt of a longer interview
  • “5–10% of AI” is an unmeasurable engagement hook — the number floats between posts

The catch

A real practice (run loops that prompt the model for you) got wrapped in a doctored quote and a shiny buzzword to sell a guide.

How to actually do it

  • Watch Cherny’s actual interview, not the clipped reel
  • For the real technique, read Anthropic’s ‘Building Effective Agents’ and look at agent loops/harnesses (LangGraph, agent SDKs) — not the ‘graph engineering’ brand
  • Ignore the ‘5–10%’ framing — it’s a hook, not a measurement

Cherny really runs loops instead of hand-prompting — that’s real. The doctored ‘graphs’ quote and the buzzword are the growth-account gloss.

Confidence
Medium
Posted by
@AnatoliKopadze (Anatoli Kopadze)

See the original claim →

We test hype for free. We build the real thing for a living.

Thirty minutes, no pitch — and you'll leave with something useful either way.

Book a call with Todd or start with the free Business Checkup →

The Verdict Weekly

Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.

© Schreier Group · schreiergroup.com · See a wild AI claim? Drop it here and we'll test it.