🔥 TOP SIGNAL

Boris Cherny (Head of Claude Code) is blunt: for the kinds of programming he does, “coding is largely solved”, and the frontier is shifting to adjacent, end-to-end agentic work (project management, paying tickets, general ops) rather than better IDE autocomplete Lenny's Podcast. In that world, throughput isn’t hypothetical: he says Anthropic saw +200% productivity per engineer (PRs)Lenny's Podcast, and Claude now reviews 100% of pull requests (with human review still in the loop) Lenny's Podcast.

🛠️ TOOLS & MODELS

• Claude Code — stability + performance signals

  • v2.1.47: long-running sessions use less memory@jarredsumner.
  • Team guidance: keep reporting issues and they’ll fix them @jarredsumner.
  • Practitioner complaint: Theo reports Claude Code has “regressed an absurd amount” with UI/feedback issues (timestamps not updating, missing “thinking,” multi-minute hangs with 0 output) @theo and suggests it “needs to be rewritten from scratch” @theo.

• Cursor — agent sandboxing shipped across desktop OSes

  • Cursor says it rolled out agent sandboxing on macOS, Linux, and Windows over the last three months @cursor_ai.
  • Mechanism: agents run freely inside a sandbox, only requesting approval when they need to step outside it @cursor_ai.
  • Implementation write-up: http://cursor.com/blog/agent-sandboxing@cursor_ai.

• OpenAI Codex — pricing/availability + compute pressure

  • @thsottiaux: Codex is included with a ChatGPT subscription (even Plus has “very generous” usage) @thsottiaux; they attribute this to gpt-5.3-codex achieving “SoTA at lower cost” @thsottiaux.
  • Same source: candidates increasingly ask how much dedicated inference compute they’ll have, and usage/user is growing faster than user count → compute could be scarce@thsottiaux.

• Gemini 3.1 Pro — dev-workflow positioning (ramping up)

  • Google Antigravity: Gemini 3.1 Pro is ramping to Google AI Ultra/Pro users, pitched around “advanced reasoning” and “long horizon planning” for dev workflows @antigravity. Details: https://antigravity.google/blog/gemini-3-1-pro-in-google-antigravity@antigravity.

• GitHub Copilot → Zed editor (GA)

  • GitHub: Copilot subscription support in Zed is generally available @github. Changelog: https://github.blog/changelog/2026-02-19-github-copilot-support-in-zed-generally-available/@github.

• Model choice drift + self-hosting pressure (reported trend)

  • Salvatore Sanfilippo says he’s seeing excellent programmers move off US models (Codex, Claude Code) toward Chinese open-weight models like Kimi 2.5 and GLM5Salvatore Sanfilippo, often via providers or by building in-house Nvidia GPU inference to avoid outages and keep sensitive data internal Salvatore Sanfilippo.
  • He frames DeepSeek v4 as a potentially major moment if it lands as SOTA (as rumors suggest), putting pressure on OpenAI/Anthropic business sustainability Salvatore Sanfilippo.

💡 WORKFLOWS & TRICKS

• “Plan mode → execute” as a default loop (Claude Code / Boris Cherny)

  1. Start the task in plan mode (he says he does this for ~80% of tasks) Lenny's Podcast.
  2. Iterate on the plan (model goes back-and-forth) Lenny's Podcast.
  3. Once the plan is good, let it execute; he’ll auto-accept edits after that Lenny's Podcast.
  • Implementation detail: plan mode is literally a prompt injection: “please don’t write any code yet” Lenny's Podcast.

• Parallel agents, but treat “state” as a first-class problem

  • Cherny: he runs ~5 agents in parallel while working (terminal/desktop/iOS) Lenny's Podcast and highlights you can run many sessions in parallel Lenny's Podcast.
  • Kent C. Dodds: similar “utter chaos” workflow—multiple projects, “a couple cloud agents” each, plus a locally guided agent @kentcdodds.
  • Failure mode (real): Simon Willison describes “parallel agent psychosis”—losing track of where a feature lives across branches/worktrees/instances @simonw.
  • Recovery trick: after hacking in /tmp and crashing, he recovered the code from ~/.claude/projects/ session logs, and Claude Code could extract and recreate the missing feature @simonw.

• Turn your feedback firehose into PRs (fast iteration loop)

  • Cherny’s pattern: point Quad/Cowork at an internal Slack feedback thread; it proposes changes and opens PRs quickly, which encourages more feedback because users feel heard Lenny's Podcast.
  • Bug-fix loop: “as long as the description is good,” he can fix a bug in minutes by delegating to Claude Lenny's Podcast.

• Token policy as a productivity lever (especially early)

  • Cherny recommends giving engineers as many tokens as possible early (even “unlimited tokens” as a perk) so they try ideas that would otherwise feel too expensive; optimize/cost-cut after an idea works Lenny's Podcast.

• Avoid over-orchestration: tools + goal > rigid workflows (model-first design principle)

  • Cherny: don’t “box the model in” with strict step-by-step workflows; give it tools + a goal and let it figure it out—he argues heavy scaffolding mattered a year ago but often isn’t necessary now Lenny's Podcast.

• “Ephemeral app” mindset + AI-native interfaces (Karpathy)

  • Karpathy built a one-off cardio experiment dashboard with Claude; it had to reverse engineer a treadmill cloud API, process/debug data, and build a web UI; he still had to chase bugs (units, calendar alignment) @karpathy.
  • His takeaway: the app-store model feels outdated for long-tail needs; instead, the industry needs AI-native sensors/actuators with agent-friendly APIs/CLIs so agents don’t have to click HTML UIs or reverse engineer services @karpathy.

• Agent “memory” ops in practice (LangSmith Agent Builder)

  • LangChain’s concrete guidance:
    • Tell your agent to remember what works@LangChain
    • Use skills to inject specialized context when needed @LangChain
    • Edit agent instructions directly when it’s faster @LangChain
  • Entry point: https://blog.langchain.com/how-to-use-memory-in-agent-builder/?utm_medium=social&utm_source=twitter&utm_campaign=q1-2026_ab-philosophy_aw@LangChain.

👤 PEOPLE TO WATCH

• Boris Cherny — production-grade Claude Code habits (plan mode, parallel sessions) + strong claims about where “after coding” goes Lenny's Podcast.
• Andrej Karpathy — high-signal framing: ephemeral bespoke apps + “AI-native CLI/API” requirements for tools and hardware vendors @karpathy.
• Simon Willison — the best micro-case study of parallel-agent failure/recovery using session logs as the source of truth @simonw +1.
• Steve Ruiz (tldraw) — pragmatic company-building: code gets easier, but alignment/positioning/communication get harder—and he’s automating the overhead away Changelog.
• Theo — sharp practitioner critique on Claude Code regressions plus continued pressure on “harness vs infra” policy differences across vendors @theo +1.
• François Chollet — frames agentic coding as ML optimization (spec/tests as constraints) and asks what the “Keras of agentic coding” will be @fchollet; @swyx suggests DSPy as the presumptive community default @swyx.

🎬 WATCH & LISTEN

1) Boris Cherny — “Plan mode” as the default starter move (~1:09:52–1:10:41)

Hook: a simple, copyable workflow: force planning first (no code), iterate the plan, then execute + auto-accept when the plan is solid Lenny's Podcast.

2) Boris Cherny — “Coding is largely solved… what’s next?” (~0:18:19–0:19:06)

Hook: his thesis on why the frontier is shifting from IDE coding to adjacent operational tasks and general automation Lenny's Podcast.

3) Steve Ruiz — daily automated release notes from landed PRs (~0:20:35–0:21:02)

Hook: treat agents like scheduled staff: every day, Claude scans the last 24h PRs and drafts “release notes we’d publish if we shipped main today” Changelog.

📊 PROJECTS & REPOS

• NanoClaw — “Clawdbot” in ~500–700 LOC TypeScript using Apple container isolation for sandboxing/security; posted as Show HN @betterhn20 +1. Repo: https://github.com/gavrielc/nanoclaw@betterhn20 • HN: https://news.ycombinator.com/item?id=46850205@betterhn20.
• Nullclaw — “fastest, smallest OpenClaw clone”: 678 KB static binary, no runtime/VM/framework overhead @jedisct1. Repo: https://github.com/nullclaw/nullclaw@jedisct1.
• tldraw agent starter kit — Cursor-like agent panel next to a canvas; cloneable starter for agent+canvas UX: https://tldraw.dev/starter-kits/agentChangelog.

Editorial take: As agents make code cheap, the new edge is orchestration discipline: plan-first loops, sandboxing, session-log recoverability, and AI-native interfaces that don’t force your agent to “be the computer.” Lenny's Podcast +3