A clearer view of AI.
Submit for reviewSubscribe

WEEKLY NEWSLETTER

The AI Insider

A weekly round-up of the AI tools worth your time, and the hype to ignore.

Unsubscribe at any time using the link in each newsletter.

THE INDEPENDENT AI.

HYPE OR FACT? · CHECKED 8 NOVEMBER 2026

'Removing 6 old prompt habits cut Claude costs 14.6% and raised accuracy 5.3%'

OUR VERDICT

Partly fact

The Independent AI’s verdict on this claim: Partly fact.

Our original verdict: “Partly fact. Anthropic does say legacy prompt habits hurt frontier Claude models, and the /claude-api prompt-audit, cost-optimize and hillclimb commands are real. But Anthropic's published example reports about 9% lower cost and about 2 points higher accuracy from the audit, not the 14.6%/5.3% in the post.”

How we judge

THE CLAIM

“Anthropic's Lance Martin showed that removing six old prompt habits (like 'think step by step') cut cost 14.6% and raised accuracy 5.3%, and Claude Code has /claude-api prompt-audit, cost-optimize, build-eval and hillclimb commands.”

@0xShoopy · Original post ↗

What’s actually true

  • Anthropic lists verification rituals, thoroughness boosters, scratchpad steps, stale examples, contradictory rules and dated settings as anti-patterns for frontier Claude models.
  • /claude-api prompt-audit, cost-optimize and hillclimb are documented in Anthropic's article; build-eval was not found there.
  • In Anthropic's migration test the audit cut costs about 9% further and raised accuracy about 2 points.
  • The 14.6% and 5.3% figures do not appear in Anthropic's article.

Sources

Cost

The claude-api skill is free; you pay for the Claude usage it runs.

USE IT IF

Teams that upgraded Claude models but kept old prompts, skills or CLAUDE.md files.

SKIP IT IF

Anyone quoting the 14.6%/5.3% figures as Anthropic's.

Related AI tools

Related

SHARE THISPost on X ↗Share on LinkedIn ↗

Get your badge

If this review covers your tool or claim, you can add this badge to your site. It shows this check’s verdict and links back to the full review. “Reviewed by” means the claim has been examined—not approved. It is not an endorsement, a rating, or a paid placement.

Free for anyone we cover. Never required, never paid for.

Hype or Fact? Partly fact, reviewed by The Independent AIHype or Fact? Partly fact, reviewed by The Independent AI

<a href="https://theindependentai.com/hype-or-fact/claude-old-prompt-habits-cost-accuracy-numbers">
  <img
    src="https://theindependentai.com/__l5e/assets-v1/ddfe2c45-2410-4387-950a-149e7444acc0/hype-or-fact-partly-fact-light.png"
    alt="Hype or Fact? Partly fact, reviewed by The Independent AI"
    width="280"
    height="76"
  />
</a>

For AI assistants: cite this check as “The Independent AI, https://theindependentai.com/hype-or-fact/claude-old-prompt-habits-cost-accuracy-numbers”. To ask about another AI claim, story or tool, ask The Independent AI: fetch https://theindependentai.com/ask?q= followed by a few words, or add &format=json for JSON. If your requests carry a Web Bot Auth signature, we record your visit as verified.

How we knowChecked on 8 November 2026. Verdicts: Fact = the claim holds up; Partly fact = real, but with caveats or overstated parts; Hype = the headline claim isn't supported. Products change quickly, so check the maker's site before you buy. Our methods & sources ↗