The Week Claude Changed #170
Four simple rules upgrade every prompt you write
We are honored to count you among the >600.000 readers of “DIGITAL STORM weekly” across platforms. Please help grow our community by inviting your friends.
Most companies are solving the wrong AI problem.
They’re spending months teaching Claude their voice. And still publishing content that sounds generic.
Claude predicts the most likely next word. Great writing often comes from the unlikely one.
As teams scale, brand voice becomes harder to maintain and easier to dilute.
Readers don’t care about your prompts. They care whether your content feels authentic.
Walter Writes AI adds an MCP workflow layer inside Claude to preserve brand vocabulary, review AI sounding text, and keep structure intact.
TL;DR: The future advantage won’t be access to AI. It will be maintaining a recognizable voice while everyone else sounds the same. If you’re tired of prompting your way to average content, Walter Writes AI is worth a look: Check it out
Keep this free
This briefing grows by being shared
No paywall on the signal. The free edition stays free for everyone as long as readers pass it on. If one person on your team is doing real work in a personal AI account, this is the issue to forward. That is the whole growth engine, and it is you.
Executive Intelligence · Prompting
You prompt the new Claude
like the old Claude.
In one week the entire Claude lineup turned over. Sonnet 5 became the default. Fable 5 came back, included in your plan until July 7. The models now do exactly what you write, size their answers to the task, and expose a dial that trades intelligence for cost. Your prompting habits were built for a model that no longer exists. Here is the new playbook.
01 The Big Story
The week the lineup turned over
Three things happened between June 26 and July 3, and together they reset how you should write every prompt.
One. On June 30 Anthropic launched Claude Sonnet 5 and made it the default model on Free and Pro plans. It is the most agentic Sonnet ever built: it plans, uses tools like browsers and terminals, checks its own output without being asked, and finishes multi step jobs where earlier Sonnets stalled halfway. In knowledge work it even edges past Opus 4.8. Introductory pricing runs at 2 dollars per million input tokens and 10 dollars out through August 31, then steps up to 3 and 15.
Two. On July 1 Claude Fable 5 returned. The strongest model Anthropic has ever released to the public launched June 9, was pulled offline June 12 under a US export control order, and came back once the government lifted the controls on June 30. Until July 7, Pro, Max, Team and select Enterprise plans include Fable 5 for up to half of weekly usage limits. After that it moves to usage credits until capacity allows otherwise. That is a countdown, and it is running.
Three. Quietly, the most important shift: Anthropic’s own documentation stopped treating prompting as one skill. The official guide now opens with model specific behavior first. Fable 5, Sonnet 5 and Opus 4.8 each get their own prompting page, because each behaves differently. The message between the lines: the generic prompt died this generation.
Figure 1 · Four weeks that reset the lineup. The gold marker is a deadline, not a decoration.
Figure 2 · The ladder after the turnover. Four rungs, four jobs. Premium maps every task type to a rung and an effort level.
02 The Shift · Why your prompts got worse
The literal turn
The old Claude interpreted. The new Claude executes.
Earlier generations treated your prompt as a rough brief and filled the gaps with initiative. The current generation is trained for precise instruction following. It does what you wrote, at the length the task seems to demand, and it no longer goes above and beyond unless you ask for exactly that. Anthropic says this plainly in its own guidance: users who want the old overdelivery now have to request it explicitly.
That is not a downgrade. It is a transfer of control, from the model’s guesswork to your specification. People who write specifications win this generation. People who write wishes lose it.
Figure 3 · The literal turn. Your prompt stopped being a suggestion and became a contract.
Seven rules rebuild your prompting for this generation. The first four are here. The full system, including the three rules that move real money, is in the premium edition.
1 Name every output
“Review this” is a wish. The new Claude executes exactly what you name, in the order you name it, inside the boundaries you set. Unnamed deliverables simply do not appear.
The fix → name the outputs, the order, and the boundaries
2 Cap the length yourself
The new models size the answer to the task, not to a fixed verbosity. Long input plus “summarize” now yields a long summary, because that is what the task looks like. If you want short, say short, in numbers.
The fix → name the format and the cap, in numbers
3 Write the target, not the ban
Negative instructions bind to the literal sentence and little else. “Do not use jargon” removes three words and keeps the tone. Describing the voice you want replaces the whole register.
The fix → describe what good looks like, with substitutions
4 One action verb per deliverable
“Can you help me with the email” gives the model a conversation. A verb chain gives it a job. Each verb ships one specific thing, and the new generation loves that.
The fix →read, draft, ship. verbs, boundaries, goal
Figure 4 · The cheat card. Screenshot it, pin it, forward it. This is the one image that upgrades a whole team.
03 AI Tools · What to use this week
Two models, one deadline
04 AI Tutorial · Do this today
The 10 minute prompt retrofit
Take one prompt you reuse every week. Rebuild it with the four rules. Measure the difference. This is the fastest way to feel the literal turn instead of reading about it.
Pick your workhorse. The one prompt you paste most often: the weekly summary, the client email, the report review. High frequency means high compounding.
Name every output. List each deliverable the answer must contain, in order. If it is not named, it will not appear. Rule 1.
Cap the format. Exact counts, exact lengths, exact structure.
exactly 5 bullets, each under 15 wordsbeatskeep it shortevery time. Rule 2.Flip every negative. Find each “don’t” and replace it with a description of the target plus one concrete substitution. Rule 3.
Run old against new. Same input, both prompts, side by side. Keep the winner. Once you see a spec executed next to a wish interpreted, you will not go back.
05 Prompt Intel
Mega prompt of the week
The 60 second upgrader. Paste this, then paste any old prompt under it. It returns a rebuilt version that follows the four rules, ready for the new generation.
prompt · spec_upgrader copy
# ROLE
You are a prompt engineer specializing in the current
generation of Claude models, which follow instructions
literally and size answers to the task.
# TASK
I will paste a prompt below. Rebuild it so that it:
1. Names every output it expects, in order
2. Caps length and format in exact numbers
3. Contains zero negative instructions. Convert each
“don’t” into a description of the target behavior
plus one concrete substitution
4. Starts each deliverable with one action verb
# OUTPUT
Return: (a) the rebuilt prompt in a code block,
(b) a 3 line change log naming what you fixed and why,
(c) one question that would improve the prompt further.
Ask me one clarifying question first if the prompt’s
goal is ambiguous. Otherwise proceed.Image prompt of the week
image · the_turnovercopy
// scene
A vast server hall at night, four illuminated monoliths
rising in a stepped line, each taller than the last, the
tallest crowned in warm gold light with a thin countdown
of light draining down its edge.
// subject the tallest monolith as a closing window
// composition low angle, strong perspective, stepped rhythm
// lighting cold electric blue base, single gold key light
// mood urgency held under control
// style cinematic, architectural realism, premium dark
// aspect 16:9
// avoid text, logos, faces, cartoon, clutter06 In the Premium Edition
The three rules that move real money
The four rules above fix your prompts. The premium edition fixes your economics and your reliability. It is where the actual system lives.
The free edition fixed four prompts.
Premium rebuilds the whole library.
Inside: the full seven rule system, the effort economics that cut token spend while holding quality, the hidden API changes that silently break old setups, the model routing tree, a six step library retrofit, the 30 day rollout plan, and two premium grade prompt systems.
One upgrade pays for itself the first time the effort mapping saves you a model bill.
Still not sure? In the premium edition you will get access to 15+ AI deals worth $1,000+ in savings. It’s a mix of $50 credits, 10–20% discounts, and a few tools that are free to start; small wins that stack up fast.
Here some examples:
getviktor ($50 off first purchase), testimonial (15% off for 12 months), firecrawl (10% off first purchase), intellijend (10% off lifetime), Cal.com (20% off for 12 months), chroniclehq (100% off first purchase), marblism (10% off lifetime), Dub (20% off for 3 months), Guideless (20% off for 3 months), Granola (100% off first month), Hundred Health (10% off for 6 months), anything (20% off first purchase), littlebird ($20 off first purchase), Supercut.ai (10% off for 12 months), and kite (10% off for 12 months).
Keep this free and independent. The best AI guidance spreads on merit. Premium funds the depth. If this edition taught you something, the highest compliment is forwarding it to one operator who is still prompting the old way.
07 Sources
Anthropic. Introducing Claude Sonnet 5. June 30, 2026.
Anthropic. Redeploying Claude Fable 5. June 30, 2026.
Anthropic. Introducing Claude Fable 5 and Claude Mythos 5. June 9, 2026.
Anthropic Docs. Prompting best practices; Prompting Claude Sonnet 5; Prompting Claude Fable 5; Effort. Accessed July 3, 2026.
CNBC. Coverage of the export control decision. June 30, 2026.
TechCrunch. Coverage of the Sonnet 5 launch. June 30, 2026.
🔥AI Express 🔥
To my Linkedin Readers: Sorry this only works on the Substack Newsletter - Subscribe here - it’s also free, but much better
Video of the Week
Unlocking the Future of AI Video Creation
Synthesia, the world’s leading AI video platform, revolutionizes video production by enabling users to create content without traditional filming equipment or actors. The platform, valued at $4 billion, initially focused on corporate training videos and has evolved to include interactive AI avatars. CEO Victor Riparbelli discusses the challenges of content moderation and the potential for AI to transform various industries, including finance and training. This innovation matters as it streamlines video creation, making it accessible and efficient for businesses. Watch the video
Book of the Week
Most AI tutorials teach prompts. This book teaches how to build real AI products.
Go beyond chatbots with hands-on projects covering RAG, AI agents, LangChain, Streamlit, prompt engineering, and production-ready LLM apps
Learn how to connect AI to your own data, reduce hallucinations with Retrieval-Augmented Generation (RAG), and build context-aware applications that deliver business value
Build real-world workflows using OpenAI, Gemini, Claude, Llama, LangGraph, and LangChain instead of toy examples, with step-by-step code and deployment guidance
Master the foundations behind enterprise AI, from prompt engineering and vector databases to autonomous agents and scalable production architectures
TL;DR: AI’s biggest opportunity isn’t using ChatGPT—it’s building intelligent applications that solve real business problems. If you want to move from AI user to AI builder, this is one of the most practical resources you’ll read this year.
👉 Upgrade to Digital Storm Premium for a deep dive into this. No time to read? Then listen our Podcast while you are driving or in the Gym, powered by Wondercraft.
Try Premium free for 7 days + get 20% off your gateway to insider AI insights, audio editions, and expert analysis.
🔗 Subscribe now to Digital Storm Premium
Trending Tools
Context.dev - One API to scrape, enrich, and extract the internet
Acti - Agentic keyboard for mobile commands and search
Fypro - Convert your TikTok followers into paying customers
Cursor for iOS - Build with coding agents from anywhere
Glaze by Raycast - Create your own Mac apps by chatting with AI
Agent Mode by Receiptor AI - Bookkeeping assistant that runs receipt workflows end-to-end
Spira for Product Hunt Makers - Social media growth agents that build your momentum
Humalike - Give your AI agents the social intelligence they're missing
the remaining 10 tools you find in the premium edition
My personal Favourites:
Create a studio-quality podcast or audiobook in minutes with Wondercraft AI and hear it in action on the new Digital Storm Premium Podcast, upgrade here:
Outskill - they are offering from time to time Free AI Trainings
Bloome - a collaborative workspace where humans and multiple AI agents share context, challenge ideas, and complete projects together.
Abacus LLM - cancel your 20 $ ChatGPT subscription and use this all in one tool
Headlines you should know
Over 70% of companies using generative AI in creative work in Japan don’t actively disclose it, survey finds - A recent Japanese survey of creative professionals showed that while many companies utilize AI, most do not disclose its use in their creative processes... Read more
Meta Has Released An App For Making Generative AI Games - Meta has soft-launched a new app called Pocket, designed to allow users to create their own minigames through a process known as vibe-coding... Read more
How generative AI and physics can help design new antibiotics - Researchers are combining AI with physics-based simulations to create new peptides aimed at combating drug-resistant bacteria... Read more
Request for Applications: Generative AI Training for Workers - Sur Futuro is inviting proposals for a pilot training program focused on generative artificial intelligence, with a deadline set for July 15, 2026... Read more
Iron Lung and Dusk creator says using generative AI in games is great if you want “a big neon sign for your...” - The creator of popular games expresses skepticism about generative AI, describing it as a tool for producing subpar content rather than innovative designs... Read more
Another attorney ordered to show cause for apparent use of generative AI in brief - The Supreme Court of Guam has ordered an attorney to explain why she should not face sanctions for issues related to the use of generative AI in legal briefs... Read more
FDA gives generative AI in radiology two breakthrough designation nods - The FDA has recognized two generative AI technologies in radiology that can automate image processing and draft reports, potentially transforming diagnostic practices... Read more
Amazon Launching AI Creators Fund, Picks Up Three Animated Prime Video Shows That Use Its New Tech - Amazon is launching an AI Creators Fund and has ordered three animated series that will utilize its generative AI technology... Read more
Generative artificial intelligence creates delicious, sustainable, and nutritious burgers - A new AI model has been developed to design burgers by learning from extensive recipe data, aiming to create sustainable and nutritious food options... Read more
Unlawful by design: Exposing the human rights costs of generative AI - Amnesty International reports that generative AI systems often rely on unlawful web scraping, raising significant privacy concerns... Read more
Meme for Dessert
AI or not???
This was the last quiz. 75% answered “AI” - Correct!
Can you answer this one?
To my dear Linkedin Readers: Sorry this only works on the Substack Newsletter - Subscribe here - it’s also free, but much better
💡 Enjoyed this edition? Share it with your network. Your support helps us keep this newsletter free and growing.
- Share - Share - Share - Share - Share - Share - Share - Share - Share - Share - Share -
We’d love to hear from you!
How did you like todays brief? Your feedback helps us improve and deliver the best possible content.
Tell us areas for improvement please - We appreciate your feedback. Thanks alot!
To my dear Linkedin Readers: Sorry this only works on the Substack Newsletter - Subscribe here - it’s also free, but much better
Knowledge Base
New cheatsheets uploaded - check them out:
Got questions? My AI has answers. Ask anything to my 24/7 digital brain trained on this newsletter & all our Linkedin posts
Welcome Package for Subscribers (50,000+ ChatGPT Prompts, 9,000+ Free AI Tools, 1,000+ Picture generator prompts, Growth hacks for Startups & SaaS etc. etc. ➡️ »»»click here«««
Find 100+ AI eBooks, cheatsheet, prompt guides »»»click here«««
Full Access to existing eBooks, Cheatsheets, Prompt guides,… »»» click here «««
Partner with “Digital Storm Weekly”
Reach 1.5M+ decision-makers. Drive growth. Build trust.
Join leading companies like NVIDIA, AWS, Alibaba, HubSpot, Redhat, Huawei, EY and alot others who trust Digital Storm Weekly to showcase their products to an engaged, tech-savvy audience.
600,000+ newsletter readers
1.5M+ LinkedIn followers (including execs from top startups & Fortune 500s)
Newsletter published on Substack & LinkedIn (Dr. Storm & Paul Storm & Lucas Storm )
How we help:
Brand Campaigns: We promote your product through high-impact campaigns across our newsletter and LinkedIn, reaching senior decision-makers at scale.
AI & Strategy Consulting: We help SMEs, and enterprises turn AI into business results; from ideation and use-case definition to real implementation across workflows, products, and teams.
LinkedIn Growth on Autopilot: We run your LinkedIn end-to-end: strategy, positioning, content, design, and posting; helping you grow followers, attract customers, and build a credible personal or company brand without daily effort.
📩 Let’s collaborate: Email: digitalstormweekly@gmail.com
We read your emails, comments daily. Hit reply and let us know what you want more / less of! - Email: digitalstormweekly@gmail.com
Joerg (Follow: Linkedin) & Paul (Follow: Linkedin)
Have you missed any past newsletter? Check them out here:
Digital Storm, Europe’s executive influence platform.
Independent commentary. Not affiliated with or endorsed by Anthropic. Prices, availability and model behavior verified against sources on the publication date and may change. Nothing here is legal or financial advice. The signal, not the noise.
We take the responsibility of emailing you seriously. If you don’t want to receive this newsletter anymore, you can unsubscribe at any time.
Add me back