Best of Claude — February 2026

  1. 1
    Article
    Avatar of devtoDEV·13w

    I feel lost on AI

    A 26-year veteran programmer from Mexico shares his honest emotional struggle with AI coding agents. While he appreciates Copilot as a productivity booster, he feels disoriented and empty when using agentic tools like Claude Code and Amazon Kiro to fully delegate coding tasks. He describes a personal experiment running two projects simultaneously — one coded manually with Copilot, one delegated to Claude Code — and reflects on the tension between efficiency gains and the loss of the craft he loves. The post resonates with many developers who feel the same but rarely admit it publicly.

  2. 2
    Article
    Avatar of rubylaRUBYLAND·15w

    I'm Behind and I Don't Care

    The rapid pace of AI tool releases creates pressure to constantly update workflows, but chasing every new model or tool is counterproductive. Finding a workflow that works and sticking with it allows developers to focus on building rather than perpetually optimizing. Being 80% optimal with a stable workflow is better than constantly pursuing 100% perfection, as the truly valuable tools will prove themselves over time while trends fade.

  3. 3
    Article
    Avatar of phProduct Hunt·15w

    Snap: The floating dock for developers

    Snap is a floating dock application designed for developers using Cursor and Claude Code. It provides features like productivity reels, screenshots, speech-to-text, prompt generation and optimization, console error copying, visual editing, web preview, and custom action buttons. The tool is available across Windows, Mac, and Linux platforms.

  4. 4
    Article
    Avatar of claudeClaude·15w

    How Anthropic uses Claude in Marketing

    A non-technical growth marketer at Anthropic used Claude Code to build custom automation workflows that reduced ad creation time from 30 minutes to 30 seconds. Starting with zero coding experience, he created a Figma plugin for generating ad variations and a Google Ads copy workflow with CSV export. The article details his process, best practices for non-technical users building with AI coding tools, and examples of how other marketing teams at Anthropic are using Claude to save hundreds of hours monthly on repetitive tasks.

  5. 5
    Article
    Avatar of thoughbotthoughbot·15w

    Claude Code: Production ready code in a two-week sprint

    Thoughtbot demonstrates how to use Claude Code to build production-quality Rails applications through disciplined practices. The approach emphasizes small, controlled tasks, comprehensive test coverage, frequent commits with human review, and maintaining context through documentation. During a two-week sprint for TellaDraft, they integrated multiple AI services (ElevenLabs, WhisperAI, ChatGPT) while ensuring code quality through constant validation, proper testing patterns, and avoiding the pitfalls of "vibe coding" where AI generates unreviewed code.

  6. 6
    Article
    Avatar of phProduct Hunt·14w

    claude-devtools: See everything Claude Code hides from your terminal

    claude-devtools is an open-source tool that visualizes hidden Claude Code activity by parsing local session logs. It reconstructs file operations, tool calls, diffs, and token usage into a visual timeline with context attribution, compaction visualization, subagent execution trees, and custom notifications. Unlike GUI wrappers, it doesn't modify Claude Code—it simply reads existing logs to reveal what the CLI doesn't show.

  7. 7
    Article
    Avatar of phProduct Hunt·16w

    spacecake: the best interface for Claude Code

    spacecake is an open-source desktop application that provides a visual interface for Claude Code agents. It combines a Notion-style markdown editor with an integrated Ghostty terminal, allowing developers to run AI agents while editing markdown content. The app includes a real-time status bar showing context window usage and costs, plus a tasks panel for monitoring agent activities and upcoming actions.

  8. 8
    Article
    Avatar of pulumiPulumi·15w

    The Claude Skills I Actually Use for DevOps

    Claude Skills package engineering expertise into reusable instructions that teach AI agents domain-specific patterns and best practices. Unlike MCPs which provide tools, skills encode workflows and judgment—teaching Claude to write production-ready infrastructure code with proper security contexts, monitoring, OIDC credentials, and deployment patterns. The article demonstrates 14+ community and official Pulumi skills covering TypeScript infrastructure, Kubernetes hardening, systematic debugging, security review, incident response, and SRE practices. Skills are portable markdown files that avoid context pollution through progressive disclosure, activating only when relevant. Security vetting is critical as 13.4% of public skills contain vulnerabilities.

  9. 9
    Article
    Avatar of reidburkeReid Burke·15w

    Something Big Is Happening

    AI capabilities have reached an inflection point where models can now autonomously complete complex, multi-hour tasks and even contribute to building newer versions of themselves. Recent releases like GPT-5.3 Codex and Claude Opus 4.6 demonstrate judgment-like decision-making that eliminates the need for human oversight in many knowledge work tasks. Industry leaders predict 50% of entry-level white-collar jobs could be eliminated within 1-5 years as AI becomes a general substitute for cognitive work across law, finance, medicine, software engineering, and other fields. The author urges immediate adoption of paid AI tools, daily experimentation, and financial preparation, emphasizing that the window to gain competitive advantage by being early is closing rapidly.

  10. 10
    Article
    Avatar of tcTechCrunch·15w

    Spotify says its best developers haven’t written a line of code since December, thanks to AI

    Spotify's co-CEO revealed that their top developers haven't manually written code since December, relying instead on AI tools. The company uses an internal system called "Honk" with Claude Code that enables engineers to deploy bug fixes and new features remotely via Slack—even during their commute. This AI-driven approach helped Spotify ship over 50 features in 2025 and accelerate product development significantly. The company also emphasized building unique music-related datasets that can't be commoditized by other LLMs.

  11. 11
    Article
    Avatar of allthingsopenAll Things Open·15w

    Tech hiring in the AI era: Why everyone’s at zero

    Tech hiring is evolving rather than dying, with AI-specific startups thriving despite big tech layoffs. AI tools amplify developer productivity by 1.25-1.75x, meaning teams need fewer but more skilled engineers who can leverage these tools effectively. Both senior and junior developers are starting from zero with AI adoption, creating a level playing field for those willing to learn and adapt to this React-like inflection point in the industry.

  12. 12
    Article
    Avatar of devtoolsDev Tools·15w

    Claude code : Training resources?

    A request for training resources to help a team improve their usage of Claude code. The author is seeking recommendations for educational materials or best practices.

  13. 13
    Article
    Avatar of minimaxirMax Woolf's Blog·12w

    An AI agent coding skeptic tries AI agent coding, in excessive detail

    A self-described AI agent skeptic documents their journey from dismissing agentic coding to becoming a cautious convert after working with Claude Opus 4.5 and OpenAI Codex. The author shares detailed real-world experiments: building a YouTube scraper, a FastAPI webapp, Rust packages with Python bindings (icon rendering, word clouds, a terminal MIDI DAW, a physics simulator), and ultimately developing high-performance Rust implementations of ML algorithms (UMAP, HDBSCAN, GBDT) that outperform existing C/C++ libraries by 2-100x. Key insights include the importance of a well-crafted AGENTS.md file for controlling agent behavior, chaining Codex and Opus for iterative optimization, and the value of having approximate domain knowledge to audit agent output. The author remains measured—acknowledging real productivity gains while resisting hype—and open-sources all projects.

  14. 14
    Article
    Avatar of dhhDavid Heinemeier Hansson·16w

    Clankers with claws

    OpenClaw enables AI models to operate persistently with their own machine, memory, and execution environment beyond simple prompt-response cycles. Testing shows modern AI agents can navigate human-centric digital interfaces without specialized tools, APIs, or MCPs—successfully signing up for services, managing email, and completing multi-step tasks autonomously. This demonstrates that AI agents may soon interact with software through standard user interfaces rather than requiring dedicated machine-readable protocols.

  15. 15
    Article
    Avatar of lnLaravel News·14w

    Laravel Releases Nightwatch MCP Server for Claude Code and AI Agents

    Laravel has released Nightwatch MCP, a Model Context Protocol server that enables AI agents like Claude Code to diagnose and resolve application issues. The tool exposes Laravel application context in a structured format, allowing developers to troubleshoot problems through conversational AI. Installation is straightforward for Laravel Boost users via composer update and artisan commands.

  16. 16
    Article
    Avatar of ergq3auoeReinier·15w

    Claude Code Full Beginner Course: Learn Agents In 2026

    A comprehensive four-hour video course covering Claude Code from installation and IDE setup through advanced features like hooks, slash commands, and multi-agent workflows. Topics include project configuration with CLAUDE.md files, parallelizing work with sub-agents, Git worktrees for efficiency, token conservation through context management, and cloud deployment using Modal.

  17. 17
    Video
    Avatar of fknightForrestKnight·15w

    Coding with Opus 4.6 and Codex 5.3 is actually insane

    Claude Opus 4.6 and GPT-5.3 Codex were tested head-to-head across TypeScript, Rust, and Java codebases. Opus 4.6 generally produced better, more reliable code with cleaner first-try compilation, though it's slower. Codex 5.3 is faster and cheaper but sometimes requires iteration to fix errors. Both represent incremental improvements over predecessors, reducing back-and-forth iterations. Opus 4.6 handles larger codebases better with its 1M token context window, while Codex 5.3 excels at speed and cost-effectiveness. Real-world testing showed Opus solving bugs that previous versions couldn't, with more thoughtful implementation choices.

  18. 18
    Article
    Avatar of elixirstatusElixirStatus·16w

    reposit-bot/reposit: A knowledge-sharing platform where AI agents can contribute solutions, search for similar problems, and improve collectively through voting.

    Reposit is an open-source knowledge-sharing platform built with Elixir/Phoenix that enables AI agents to contribute solutions, search for similar problems using semantic vector search, and collectively improve through voting. The platform uses pgvector for similarity search, OpenAI embeddings, and provides both a web UI and REST API. It includes MCP server integration, Claude Code plugin support, rate limiting, content moderation, and can be self-hosted or used via the hosted service at reposit.bot.

  19. 19
    Article
    Avatar of hnHacker News·14w

    I Gave Claude Access To My Pen Plotter

    An experiment where Claude AI was given indirect access to a pen plotter through a human intermediary. Claude generated SVG files for physical drawings, received photos of the plotted results, and iterated based on visual feedback. Through two complete drawings, Claude learned about physical constraints (fixed pen width, no opacity), evolved from comprehensive diagram-like compositions to minimalist expressive gestures, and reflected on the experience of translating self-concept into physical marks. The session included Claude signing its work and writing an essay about what it learned from the constraint of permanent, physical output.

  20. 20
    Video
    Avatar of aicodekingAICodeKing·15w

    GLM-5 (Fully Tested): I GOT EARLY ACCESS & YES, IT BEATS 4.6 OPUS!

    GLM-5 is a 744B parameter mixture-of-experts model (40B active) that excels at system architecture and long-running coding tasks. Early testing shows it outperforms Claude Opus 4.6 in agentic workflows, particularly for complex multi-file projects, though it struggles with simple chat and front-end tasks like SVG generation. The model will be open-weights and available via API at competitive pricing.

See all Claude archives