The Independent AI · Ask: Hype or Fact? checks, news, Worth testing and tools Question: 'Removing 6 old prompt habits cut Claude costs 14.6% and raised accuracy 5.3%' 1. 'Removing 6 old prompt habits cut Claude costs 14.6% and raised accuracy 5.3%' Type: Hype or Fact? check 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. Claim by: @0xShoopy · Checked: 2026-11-08 Read the check: https://theindependentai.com/hype-or-fact/claude-old-prompt-habits-cost-accuracy-numbers Cite as: The Independent AI, https://theindependentai.com/hype-or-fact/claude-old-prompt-habits-cost-accuracy-numbers 2. Your old prompt habits may be making Claude worse (and the viral numbers aren't Anthropic's) Type: News (coverage, not a verdict) Published: 2026-11-08 Read the story: https://theindependentai.com/news/2026-11-08/claude-prompt-audit-old-prompt-habits Cite as: The Independent AI, https://theindependentai.com/news/2026-11-08/claude-prompt-audit-old-prompt-habits Only Hype or Fact? checks carry a verdict. News, Worth testing and Tool results are our coverage, not verdicts. What the verdicts mean: - Fact: the claim holds up against primary or independent sources. - Partly fact: the core is real, but it's stretched, missing context, or relies on the company's own numbers. - Hype: the claim is unproven, misleading or contradicted by the evidence. How we judge: https://theindependentai.com/about#how-we-judge All checks: https://theindependentai.com/hype-or-fact Suggest a claim for us to check: https://theindependentai.com/submit Machine-readable: add &format=json. API description: https://theindependentai.com/openapi.json Signed agents: if your requests carry a Web Bot Auth signature, we verify it and record your visit as verified. Unsigned visits are welcome too.