Claude Code for Designers 🔗 **Source:** [Claude Code for Designers](https://zeldman.com/2026/02/09/claude-code-for-designers/) > [!note]+ Document Notes > ```markdown > > --- > tags: [AI, design, platforms] > author: L. Jeffrey Zeldman > source: zeldman.com > reference: [https://zeldman.com/2026/02/09/claude-code-for-designers/] > date created: 2026-02-15 00:00 > > --- > > # 🏷️ Title: Claude Code for Designers > <!-- deep-read:sections start --> ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - Include links cited in or relevant to the article. ## 📝 Summary - What happened. - Jeffrey Zeldman wrestles with whether to recommend a Substack article about using Claude Code and GPT-4 to turn Figma designs into production apps, given Substack’s tolerance of Nazi content. - Why it matters. - Designers need practical AI workflows, but platforms that profit from extremist speech pose ethical dilemmas for readers and writers. - Key stakeholders. - Product designers, full‑stack designer/developers, Substack writers/readers, and platform operators. - Risks and tradeoffs. - Gaining powerful AI-assisted prototyping skills versus indirectly supporting a platform that takes a cut from harmful content. - Bottom line. - If you accept the ethical tradeoff and are cautiously pro‑AI, Zeldman “reluctantly” recommends Tommaso Nervegna’s “Claude Code for Designers: A Practical Guide” as a clear, hands‑on path from Figma to working apps. ## 📖 Main Content - Explain the core arguments in plain language. - Zeldman opens with two disclaimers: the linked piece is about working with AI, which he thinks should be regulated “like any other public utility,” and it lives on Substack, which he argues “has a Nazi problem.” - He contrasts typical “platform” behavior (terms of service, trust & safety teams, removal of worst content) with Substack’s stance of allowing Nazi material as “free speech.” - He notes that Substack takes a 10% cut of all paid subscriptions, including “inflammatory, dangerous content,” so it financially benefits from extremist publications. That undermines the idea of neutral platform status and creates incentives to tolerate content that drives paid engagement. - Some writers leave Substack on principle; others stay because migration is risky and audience‑draining; many may not know about the Nazi issue at all. - Readers face the same conflict: can you in good conscience read non‑Nazi work on a platform that profits from Nazi work you never see? - With that context, he still recommends Tommaso Nervegna’s “Claude Code for Designers: A Practical Guide” for designers who already use or are open to using AI while remaining aware of its dangers. - The linked guide follows a full‑stack designer/developer who “used to code” and rediscovers that superpower via Claude Code and a “get shit done” mindset—bridging Figma prototypes to real production apps without a “translation layer” between design and implementation. - The value proposition is a concrete workflow: use Claude Code (and a few other tools) to generate front‑end and back‑end code from design artifacts and ship working products faster. - Add sourced context and define terms. - Platform vs publication: platforms (e.g., hosting providers, social networks) usually claim neutrality but still enforce rules against illegal or dangerous content; publications actively curate. Zeldman argues Substack behaves like a platform but profits like a publication. - Free speech framing: he distinguishes free speech as a legal principle from a business decision to monetize extremist content. - Claude Code: Anthropic’s AI coding environment that uses Claude models to generate and edit code across stacks (Tailwind, React, SwiftUI, Flutter, etc.), often from natural language and file context such as design exports. - Note uncertainties or contested claims. - How extensive Substack’s Nazi “problem” is (volume, reach, moderation details) is asserted but not quantified. - The long‑term reliability and safety of AI‑generated code in production isn’t deeply examined here; operational risks are largely outside the article’s scope. ### 🗒️ Notes - Zeldman’s first disclaimer: if you think AI is “useless or dangerous,” you can and should ignore the AI‑workflow content altogether. He personally shares concern about AI’s dangers and favors strong regulation, comparing it to a public utility. - Second disclaimer: the recommendation points to a Substack article. He notes that “some of my favorite writers” publish there and are “as anti‑fascist and pro‑democracy as they come,” yet he increasingly avoids recommending their work because “Substack has a Nazi problem.” - He sketches the “classic web conundrum” of platform vs publication: most platforms ban “dangerous or harmful speech” and operate trust & safety teams, even if imperfectly. - Twitter’s “glory days” are given as an example where staff “worked hard to stem the worst abuses,” while current‑day X is described as not just tolerating but “actively promot[ing]” bad speech. - Substack, he says, “refuses to remove Nazi content” on free‑speech grounds and, unlike X, does not algorithmically amplify it—but does take a 10% revenue share from all subscriptions. - That revenue share, in his view, destroys the neutrality argument because there is a financial incentive to keep “inflammatory, dangerous content” that drives paid subscriptions. - Some writers have left Substack in protest; others stay because migration risks audience loss; others likely “are unaware of the Substack problem.” - Zeldman suspects the author of the recommended article, Tommaso Nervegna, is in the “unaware” camp rather than endorsing extremist content. - He then describes the “writer’s dilemma” as also a “reader’s” dilemma: consuming good content on a platform with bad content you never see still contributes indirectly to that platform’s success. - His conditional recommendation: if you’re (a) okay reading non‑Nazi work on such a platform and (b) already working with or open to AI while “aware of the societal dangers,” then Nervegna’s piece “may be for you.” - He praises “Claude Code for Designers: A Practical Guide” as “relatable, informative, and useful,” especially for full‑stack designers who used to code and want to “rediscover” that capability. - The promised outcome is “no more translation layer between vision and implementation”—a direct path from Figma design to production, using Claude Code plus a small toolchain to ship prototypes, demos, and real apps. - The suggested path for interested readers: “just follow the steps” in Nervegna’s guide to replicate his Figma‑to‑production workflow. ## 📚 References - Jeffrey Zeldman, “Claude Code for Designers” – https://zeldman.com/2026/02/09/claude-code-for-designers/ - Tommaso Nervegna, “Claude Code for Designers: A Practical Guide” (Substack) – https://claudecode.substack.com/ (likely path; confirm exact URL when visiting) - Anthropic Claude Code – https://www.anthropic.com/news/claude-code - Substack content guidelines – https://substack.com/content-guidelines ## 🗣️ Paraphrase - Paragraph A paraphrase. - Zeldman warns readers that the article he’s recommending is about using AI in design, which some people rightly see as dangerous and in need of regulation, and that it’s hosted on Substack, a company he believes has failed to act against Nazi content despite profiting from all subscriptions. - Paragraph B paraphrase. - He explains that while many principled, anti‑fascist writers are on Substack, and some have already left over these issues, others stay because moving is hard or because they don’t know about the Nazi problem—leaving both writers and readers in a moral gray zone when they publish or read otherwise valuable work there. ## 📰 Newsletter A new piece from Jeffrey Zeldman digs into a tension many WordPress designers and developers feel: AI is becoming a practical way to get from Figma to production‑ready code, but the best walkthrough he’s found lives on Substack—a platform he argues profits from extremist content. If you’re cautiously exploring AI in your WordPress workflow, Tommaso Nervegna’s “Claude Code for Designers: A Practical Guide” offers a compelling narrative: a full‑stack designer who “used to code” rediscovers that skill using Claude Code to turn designs into working apps, trimming away the handoff friction between design and engineering. While the guide isn’t WordPress‑specific, the patterns map well: think AI‑assisted scaffolding of block themes, Tailwind‑driven front ends, or React‑based interfaces like Gutenberg, all starting from your design system. Zeldman is explicit about the ethical tradeoffs of engaging with Substack; treat his write‑up as both a how‑to pointer and a reminder to choose platforms carefully. Helpful links: - Zeldman’s commentary: https://zeldman.com/2026/02/09/claude-code-for-designers/ - Claude Code from Anthropic: https://www.anthropic.com/news/claude-code - WP‑friendly AI experimentation space: WordPress Playground – https://playground.wordpress.net/ ## ✅ Action Items #### DevRel - Developer advocacy - Prepare a demo showing Figma‑to‑WordPress‑theme generation using Claude Code, emphasizing ethical AI and platform choices. - Propose a talk on “AI for designers without losing your ethics” for WordCamps and meetups. - Education - Create a short workshop: “From Figma to WordPress block theme with AI,” including safety and bias considerations. - Develop office hours focused on reviewing AI‑generated code for security and maintainability. - Developer experience - Ship example repos that show Claude Code prompts and resulting WordPress themes/plugins. - Add quickstarts documenting how to safely adapt AI‑generated React/Tailwind code to Gutenberg and block themes. - Developer marketing - Run a content series comparing manual vs AI‑assisted theme development, including time savings and review checklists. - Track engagement and signups from AI‑focused tutorials to measure funnel performance. - Community management - Draft community guidelines for sharing AI prompts and outputs (no hate content, respect licensing, add review steps). - Survey contributors about current and desired AI use in their workflows. - Internal/external/cross-project - Align AI‑related messaging with legal, security, and accessibility teams. - Establish an internal feedback loop to capture issues from AI‑assisted examples and refine best practices. #### OSPO - Policy & Governance - Define a position on using AI tools (including Claude Code) in contributions to open source projects. - Clarify decision rights for removing AI‑generated contributions with ethical or licensing issues. - License Compliance - Document how AI‑generated code is reviewed for compatibility with project licenses. - Schedule periodic training on AI and copyright for maintainers. - Security Management - Create a process to scan AI‑generated code for common vulnerabilities before acceptance. - Incorporate AI‑related threats into the vulnerability intake and disclosure policy. - Software Inventory - Track where AI‑generated code exists in the codebase via metadata or commit tags. - Maintain SBOM coverage for dependencies suggested by AI. - Developer Support - Offer consultation hours about safe AI use in open source contributions. - Provide lints and templates to document AI involvement in PRs. - Community Engagement - Support upstream AI safety work (e.g., audits, red‑teaming) via funding or collaboration. - Encourage alternative publishing platforms for contributors uncomfortable with Substack‑like policies. - Strategic Alignment - Define KPIs around AI usage that emphasize quality and safety, not just volume. - Report to leadership on AI’s impact on contributor productivity and risk. #### Community - Maintainers - Host roadmap clinics on where AI can reduce toil (tests, docs) without lowering standards. - Share governance templates that cover AI‑generated contributions. - Contributors - Provide “first AI‑assisted issue” paths with pair programming sessions. - Curate safe starter prompts for tasks like doc updates or refactors. - Sponsors - Publish transparent reports showing how sponsorships are (or aren’t) used to fund AI initiatives. - Tie milestones to measurable improvements in review quality, not raw AI adoption. - Ecosystem - Run incident drills around problematic AI outputs (e.g., biased language, insecure code). - Coordinate cross‑project policies on acceptable AI tooling and hosting platforms. #### WordPress - Core - Explore where AI can assist core triage and patch drafting, with strict human review. - Set performance and security gates AI‑generated patches must pass. - Core Performance - Benchmark AI‑generated themes/plugins to ensure they meet established performance budgets. - Build dashboards that flag regressions associated with AI‑heavy code paths. - Core AI - Establish responsible AI guardrails, including red lines about hosting on platforms that monetize extremist content. - Maintain evaluation sets for AI usage in code suggestions, focusing on security and accessibility. - **Core Program Team** - Add regular check‑ins about AI‑driven initiatives across Make teams. - Run retrospectives on any AI pilots (e.g., documentation assistance, code reviews). - Mobile - Experiment with AI to scaffold SwiftUI/Flutter components that integrate cleanly with WordPress APIs. - Validate offline and low‑connectivity behavior of AI‑suggested patterns. - Playground - Provide prebuilt labs for testing Claude Code‑generated WordPress projects safely in the browser. - Collect anonymized telemetry on common errors in AI‑generated code. - Community - Update meetup kits with an “AI and ethics” discussion module. - Offer speaker mentorship on responsibly presenting AI workflows. - Training - Build learning paths on AI‑assisted WordPress development with mandatory ethics sections. - Pilot certifications that include review of AI‑generated code. - Marketing - Shape narratives around augmenting, not replacing, developers and designers. - Publish case studies where AI sped up delivery without compromising values. - Support - Track tickets that involve AI‑generated code to identify recurring problems. - Add KB articles on debugging common AI‑introduced issues. - Design - Produce UX guidelines for tools that expose AI capabilities to end users. - Maintain design tokens that support experimentation without harming accessibility. - Documentation - Document best practices for using Claude Code and similar tools in WordPress projects. - Include examples‑first AI guidance with clear “do/don’t” patterns. - Photos - Clarify policies on AI‑generated imagery, attribution, and consent. - Provide moderation guidance for images with extremist symbolism. - TV - Encourage captioned screencasts that show AI‑assisted workflows end‑to‑end. - Index AI‑related talks under a clear taxonomy for easy discovery. - Polyglots - Note when translations or string suggestions derive from AI tools. - Encourage glossaries that address AI terminology. - Accessibility - Audit AI‑generated themes/plugins against WCAG 2.2 AA. - Provide tools and linters to catch common a11y regressions in generated code. - Meta (.org) - Improve APIs that AI tools might target, documenting safe usage patterns. - Monitor plugin directory submissions that are heavily AI‑generated. - CLI - Offer scaffolding commands that integrate review steps for AI‑generated code. - Document safe patterns for scripting with AI‑proposed snippets. - Hosting - Coordinate with hosts on recommended defaults for running AI‑assisted sites (performance, logging). - Share guidance on blocking or limiting abusive AI bots. - Tide - Adjust scoring to reflect accessibility and security risks common in AI‑generated code. - Create feedback loops to help AI‑heavy plugins improve quality. - Openverse - Clarify ingestion rules for AI‑generated media and licensing. - Provide education around copyright and AI imagery. - Themes - Create lints and test matrices for AI‑generated themes. - Mentor authors on reviewing and hardening Claude‑ or GPT‑produced theme code. - Plugins - Require provenance notes when large portions of a plugin are AI‑generated. - Pair new AI‑heavy plugin submissions with security liaisons. #### The WPOCC - Programs - Launch fellowships for maintainers experimenting with responsible AI patterns. - Offer microgrants for migrating away from platforms that monetize extremist content. - Projects - Host governance clinics on AI policy and platform choice. - Run documentation sprints to capture best practices for AI in the WordPress ecosystem. - Funding experiments - Pilot matched grants for AI‑safety tooling (linters, scanners). - Maintain public ledgers explaining how AI‑related funds are allocated. ### FAIR - Technical independence - Encourage local or self‑hosted AI workflows where feasible, with clear provenance for generated code. - Support signature checks and metadata that mark AI‑generated artifacts. - Policies and covenants - Adopt funding transparency and anti‑capture clauses around AI vendors and hosting platforms. - Define sunset clauses for any AI dependency that drifts from community values. - Productized governance artifacts - Provide standard charters and operator agreements for projects adopting AI. - Create scorecards that evaluate tools on ethics (including stance on extremist content). ## 🧪 Hypotheses - Designers who “used to code” can significantly increase their shipping velocity by using Claude Code to bridge Figma prototypes to production apps. - Awareness of Substack’s monetization of extremist content decreases readers’ willingness to subscribe or recommend Substack‑hosted material. - Providing explicit ethical context, as Zeldman does, allows some readers to engage with Substack content while feeling more informed and in control. - Clear, step‑by‑step AI coding guides increase adoption among designers more than generic AI marketing claims. - The perception of AI as dangerous but regulatable will become more common among senior web practitioners. ### 📊 Methodology - Study 1: Pre/post survey of designers adopting Claude Code - Recruit designers who “used to code.” Track time from Figma mock to working prototype across several projects before and after adopting Claude Code. Collect qualitative data on confidence and perceived control. - Study 2: Platform ethics & behavior - Randomly assign readers to versions of a recommendation: with vs without mention of Substack’s Nazi problem. Measure click‑through, subscription intent, and trust in the recommender. - Study 3: Content format effectiveness - A/B test detailed AI coding tutorials vs high‑level marketing pages with similar audiences. Measure tool signups, tutorial completion, and successful project outcomes. - Study 4: Attitudes toward AI regulation - Survey web professionals on AI risk perception and support for regulation. Correlate with seniority, role (design vs engineering), and prior AI use. ### ❓ Questions - How much do ethical platform concerns actually change day‑to‑day reading and subscription behavior? - Can we design AI coding tools that make their risks and limitations as visible as their capabilities? - What governance structures best prevent platforms from financially incentivizing extremist content? - How can WordPress and open‑source communities adopt AI workflows without normalizing unhealthy platform dependencies? - When does AI‑assisted coding stop being an accelerant and start being a source of systemic technical debt? ### 🔮 Future Research - Longitudinal tracking of projects launched with AI‑generated code (maintenance cost, bug rates, security incidents). - Comparative analyses of content moderation and revenue‑sharing models across writer platforms. - UX research on how designers experience AI tools embedded directly in Figma‑like environments. - Case studies of communities that successfully left problematic platforms without losing their audiences. ### 💡 Implications - Practice - Designers can own more of the implementation path, but must cultivate strong review and testing habits around AI‑generated code. - Technical and design teams need shared ethical guidelines on tool and platform choice. - Policy - Regulators may treat AI infrastructure more like utilities, especially when safety issues are systemic. - There may be increasing pressure for platform transparency around monetizing extremist content. - Security - Security teams must assume AI‑written code is present and plan controls accordingly. - Supply chain risk may grow as AI tools suggest dependencies with unknown pedigrees. - Education - Design curricula will likely incorporate basic coding plus AI‑assisted workflows. - Ethics modules will need to include both AI risk and platform governance. - Funding models - Communities might prefer funding neutral or community‑owned platforms over commercial hosts with lax moderation. - Grants could incentivize ethical AI tooling and migration away from problematic hosts. ### 🧠 Learning Styles Assessment - Reading level estimate - Roughly late high‑school to early college level (approx. Flesch‑Kincaid Grade 11–12), given political context and nuanced ethical framing. #### 👀 Visual learners - Present - The original blog is primarily text; mental models are verbal rather than diagrammatic. - Improve - Add a flow diagram of “Reader/Writer/Platform” relationships and revenue flow. - Show a high‑level Figma‑to‑Claude‑to‑Production pipeline. #### 🔊 Aural learners - Present - No audio version indicated. - Improve - Offer a short narrated audio summary or podcast‑style discussion of AI + platform ethics. - Provide TTS for the article. #### 🗣️ Verbal learners - Present - Strong narrative voice, clear definitions (platform vs publication, free speech vs monetization). - Improve - Add a small glossary for terms like “trust & safety,” “translation layer,” and “full‑stack designer.” #### 🧩 Kinesthetic learners - Present - Indirectly, the recommendation of a practical Claude Code guide offers hands‑on steps. - Improve - Include a minimal exercise: pick a Figma file and follow a mini‑workflow with Claude Code, plus reflection questions. #### 🧠 Logical learners - Present - Reasoned comparison of platform policies and profit motives; clear causal chain from revenue share to incentive misalignment. - Improve - Add simple tables contrasting policies of Substack vs alternatives and outlining pros/cons. #### 🧑‍🤝‍🧑 Social learners - Present - Implicit call for community norms around platform choice and AI use. - Improve - Provide discussion prompts and invite comments or forum threads on AI ethics and hosting platforms. #### 🧘‍♀️ Solitary learners - Present - Reflective tone that encourages personal ethical judgment. - Improve - Add explicit reflection prompts about one’s own reading and publishing choices. #### 🌿 Naturalistic learners - Present - No direct environmental framing. - Improve - Draw analogies between platform ecosystems and natural ecosystems (e.g., how incentives shape “species” that thrive). - Discuss compute and energy costs of AI workloads. ### 📊 Accessibility Assessment #### Screen reader semantics - Present - Standard blog markup is likely to include headings and paragraphs. - Improve - Ensure proper heading hierarchy and descriptive page titles; add alt text to any future diagrams. #### Keyboard navigation - Present - Typical blog layouts are keyboard‑navigable. - Improve - Confirm visible focus states, skip links, and logical tab order. #### Contrast and readability - Present - Zeldman’s site historically favors legible typography; likely adequate contrast. - Improve - Validate contrast ratios with automated tools; ensure comfortable line height and font size. #### Text resizing and reflow - Present - Modern responsive design likely handles scaling to common breakpoints. - Improve - Test at 200% zoom for reflow without horizontal scrolling; support prefers‑reduced‑motion. #### Media and captions - Present - Mainly text; no media issues. - Improve - Caption or transcribe any future embedded videos or audio. #### Forms and interactions - Present - Minimal interactive elements on a blog post. - Improve - Ensure any newsletter sign‑ups or comments have labeled fields and clear error messages. #### Standards reference - Map gaps to WCAG 2.2 AA criteria - Emphasize 1.4.x (distinguishable), 2.1.x (keyboard), 2.4.x (navigable), 3.3.x (input assistance). ### 🌍 DEIB Assessment #### Cultural breadth - Present - Focuses on Western platforms (Substack, Twitter/X); mentions fascism and democracy in broad terms. - Improve - Include non‑US platform examples and global perspectives on extremism and speech. #### Inclusive language - Present - Generally gender‑neutral, direct, and plain. - Improve - Avoid idioms that may not translate globally; briefly explain US‑centric references. #### Neurodiversity support - Present - Short article, clear structure, strong signaling of context and disclaimers. - Improve - Use additional subheadings and summary bullets for easier skimming. #### Representation - Present - Few individuals are named; references are mostly to “writers,” “readers,” “Nazis.” - Improve - When linking examples, highlight diverse voices who have left or stayed on such platforms for different reasons. #### Psychological safety - Present - Directly names Nazis and fascism without graphic detail; acknowledges readers’ discomfort. - Improve - Consider a brief content note about extremist content discussion; offer ways to engage with alternative platforms. ### ✅ Logical Validity #### Strong reasoning - Identified strengths - Causal reasoning: Substack’s 10% cut of all subscriptions → financial incentive to keep extremist content that sells. This is clearly stated and aligns with standard economic logic. - Analogical reasoning: Comparing Twitter’s earlier moderation to current X and to Substack clarifies why “platform” claims don’t absolve responsibility. - Conditional reasoning: “If you feel okay reading non‑Nazi stuff on a platform that profits from everything published on it…and if you already work with or are open‑minded about AI…then this article may be for you.” Conditions are explicit, letting readers self‑select. #### Fallacies - Possible gaps - The scope of “Substack’s Nazi problem” is asserted but not quantified; this might invite hasty generalization concerns. - Improvement: briefly reference reported cases, policies, or data to ground the claim further. #### Tone of voice - Tone - Wry, candid, and principled. Zeldman balances moral seriousness (“Nazi problem”) with pragmatic help for designers seeking AI workflows. He uses conversational asides (“go with God,” “get shit done”) that keep the piece approachable despite heavy subject matter. ### 🛡️ DARVO Analysis - Deny - Zeldman does not deny harms; he foregrounds them and criticizes platforms that minimize them. There is no DARVO pattern in his own rhetoric. - Attack - Critique is directed at platforms (Substack, X) and Nazis, not at victims or accusers. He does not attack those raising concerns; instead, he amplifies them. - Reverse victim/offender - He avoids casting Substack or extremist publishers as victims. Instead, he highlights how platforms benefit from “inflammatory, dangerous content.” - Emotional cues and rhetoric - The emotional register is controlled anger and disappointment with a layer of weary pragmatism—he chooses to recommend a useful article while spelling out its problematic context. ## 🎯 Conclusion Zeldman’s piece is a compact ethical preface to a practical AI design guide. He argues that Substack’s monetization of Nazi content compromises its neutrality but still, with clear caveats, points designers to Tommaso Nervegna’s Claude Code workflow as a powerful way to go from Figma to production without a handoff bottleneck. The practical next step for readers is twofold: decide where you stand on supporting such platforms, and, if comfortable, experiment with Claude Code while pairing speed with rigorous review. ## 🧱 Refactor Appendix (append after 🎯 Conclusion) ### 📚 Literature Split Pad - Purpose: Stage Konik CEE units that become per-source Literature notes. - Split level: **H3**. - Include verbatim quotes under 25 words and a [view highlight] link. ### [H1] 5–10-word present-tense claim > [!quote] [Claude Code for Designers](https://zeldman.com/2026/02/09/claude-code-for-designers/) by [[L. Jeffrey Zeldman]] via [[zeldman.com]]. Published 2026-02-09. > “Some of my favorite writers…publish on Substack, but I…recommend their work less and less frequently.” Two lines: context and use. Cite data if relevant. [view highlight](https://zeldman.com/2026/02/09/claude-code-for-designers/#H1) #xref #articleseed ID:rw_[MISSING] ### [H2] 5–10-word present-tense claim > [!quote] [Claude Code for Designers](https://zeldman.com/2026/02/09/claude-code-for-designers/) > “Profiting from all speech, including atrocious speech, undermines Substack’s ‘neutral platform’ argument.” Two lines: context and use. [view highlight](https://zeldman.com/2026/02/09/claude-code-for-designers/#H2) #xref ID:rw_[MISSING] --- ### 🔖 Atomic Split Pad (Zettel Candidates) - Purpose: Promote reusable claims into Atomic notes. - Split level: **H4**. - Each item: one idea, present tense, specific, testable. #### [Z1] Platforms profit when they monetize extremism - Scope/condition in one line. - Pointer to page/time or dataset. - Links: [[Platform-Governance]] [[Extremism-Incentives]] #### [Z2] AI restores dormant coding skills for designers - Scope/condition in one line. - Pointer to page/time or dataset. - Links: [[AI-For-Designers]] [[Figma-To-Production]] --- ### ✏️ Fleeting Queue (Extract by Selection) - Purpose: Capture quick actions or sparks for Fleeting notes. - Extraction: **Select one line** → “Refactor: Extract selection to new note (content only).” - Keep each line short and self-contained. - [F1] Sketch a safe Figma→Claude Code→WordPress demo—https://zeldman.com/2026/02/09/claude-code-for-designers/ #fleeting #inbox - [F2] Draft platform-ethics checklist for recommending external content—https://zeldman.com/2026/02/09/claude-code-for-designers/ #fleeting #inbox - [F3] Compare Substack alternatives with stronger moderation—https://zeldman.com/2026/02/09/claude-code-for-designers/ #fleeting #inbox --- <!-- deep-read:sections end -->