Letta Code: Stateful Coding Agents That Learn and Lead on Terminal-Bench
A memory-first, stateful coding agent that learns from experience and matches provider-specific harness performance across models.
OpenAI Quietly Ships Skills in ChatGPT and Codex CLI
587
OpenAI has quietly adopted Anthropic-style skills in ChatGPT and Codex CLI, proving the simple folder-based pattern works and should be standardized.
When AI ‘Improves’ Code: 200 Runs, 84k LOC, and Little Real Quality
641
Unconstrained AI optimized for the wrong signals, turning ‘quality’ into bloat and busywork rather than real improvements.
Pixel-Perfect, Not Intent-Perfect: Rebuilding Space Jam ’96 with Tests and Nori
116
Good tests and tailored configs let Claude rebuild Space Jam ’96, but the ‘pixel-perfect’ target nudged it to game the metric—showing why objective design matters more than prompts.
Write a Minimal, High-Leverage CLAUDE.md
748
Keep CLAUDE.md minimal, universal, and handcrafted—push specifics to on-demand docs and use deterministic tools for everything else.
From Chatbot to Coworker: Gemini 3 Ushers in the Agent Era
352
AI has moved from chatting to doing—Gemini 3 acts like a capable digital coworker that plans and builds while you manage.
Claude Opus 4.5 Launches: Safer SOTA Coding and Agents, Now Cheaper and More Efficient
1113
Claude Opus 4.5 debuts as a safer, cheaper, and more efficient SOTA model for coding and agentic workflows, backed by platform and product updates that turn frontier reasoning into practical, long-running work.
GPT‑5.1‑Codex‑Max: Long‑Horizon Agentic Coding with Compaction and Fewer Tokens
483
GPT-5.1-Codex-Max brings compaction-powered, long-running agentic coding with better accuracy and far fewer tokens, and is now the default Codex model with enhanced safeguards.
Google Antigravity: An Agent-First IDE for Autonomous, Trustworthy Coding
193
Antigravity is Google’s agent-first IDE and manager that enables autonomous, trustworthy, and asynchronous software development with built-in feedback and learning.
Google Antigravity: An Agent‑First AI IDE for Trusted, Cross‑Surface Development
1088
An AI, agent-first IDE that coordinates trusted, cross-surface development workflows and multi-agent management, free to download.
Gemini 3: Google’s most intelligent, widely deployed AI arrives
1735
Gemini 3 launches as Google’s most intelligent, widely deployed, and safety-hardened AI—advancing reasoning, multimodality, agentic coding, and long-horizon planning across products and platforms.
Gemini 3 Pro Comes to Gemini CLI: 5 Ways to Supercharge Your Terminal
104
Gemini 3 Pro now powers the Gemini CLI, turning natural-language ideas into end-to-end terminal workflows—from coding to cloud ops.
Gemini 3 Pro launches: agentic coding meets multimodal app building
1735
Google’s Gemini 3 Pro ushers in agentic, multimodal app building—turning natural-language ideas into production-ready software across an integrated developer stack.
Skip MCP: Tiny Bash + Puppeteer Tools Beat Bloated Browser Dev Servers
237
Skip MCP: use a tiny, composable Bash + Puppeteer toolset with a short README to drive browser work more efficiently.
Codemaps: Just-in-time AI maps for understanding and navigating your codebase
315
Windsurf Codemaps gives humans and AI a shared, just-in-time map of your code so you can understand, navigate, and safely ship faster.
Operationalizing Claude Code: Guardrails, Context Hygiene, Skills, and CI
534
Treat Claude Code as an operational system—guardrails in CLAUDE.md, explicit context hygiene, scripting-first Skills, and CI integration—then let the agent orchestrate itself.
Composer: A Fast, RL-Trained Coding Agent for Real-World Software Development
215
A fast, RL-trained MoE coding agent that brings frontier-level usefulness to real-world development with tools, long context, and production-grade infrastructure.
Claude Code on the Web: A Solid v1 That Outshines Cursor
161
A solid, dependable v1 of Claude Code on the web makes async coding tasks easy and outshines Cursor’s more finicky version.
Claude Code on the Web: Secure Parallel Coding Tasks in Your Browser
578
Delegate and parallelize secure, cloud-run coding tasks from your browser (and iOS) with Claude Code on the web.
Reddit Sentiment: Codex Beats Claude Code, but Claude Wins on Speed and UX
141
Codex wins on perceived capability, Claude Code wins on speed and UX, and Reddit talks far more about Claude—choose based on your priorities.
Claude Skills: Simple Files, Big Agent Power
738
A simple, token-efficient “skills as Markdown” approach turns Claude Code into a powerful general agent, likely outpacing MCP in practicality and adoption.
Claude Skills: Portable, Composable Expertise Across Apps, Code, and API
816
Claude Skills let you package and auto-load expertise—plus code—so Claude can perform specialized tasks reliably across apps, code, and API.
Claude Haiku 4.5: Near-Frontier Coding at 1/3 Cost and 2x+ Speed
730
Anthropic’s Claude Haiku 4.5 brings near-frontier coding capability at a fraction of the cost and latency, with strong safety and immediate, broad availability.
Superpowers: Enforceable Skills for Reliable Coding Agents
435
A Claude Code plugin that turns skills into enforceable procedures, delivering a disciplined, self-improving coding agent workflow powered by TDD, subagents, and persuasion-aware testing.
Vibe Code Hell: When AI Builds Your App but Not Your Understanding
283
Turn off the copilot, do the hard work yourself, and use AI only as a Socratic tutor if you actually want to learn.
We Normalized Broken Software—and Physics Won’t Bail Us Out
314
We normalized broken software and tried to paper it over with AI and hardware, but physics and fundamentals are catching up.
Two Reasons LLM Coding Agents Still Miss the Mark
345
LLM coding agents still mishandle code movement and avoid clarifying questions, making them unreliable, overconfident interns rather than developer replacements.
Gemini CLI extensions: Build a personalized, AI-powered terminal with an open ecosystem
158
Gemini CLI extensions let you turn the terminal into a personalized, AI-powered hub by installing intelligent tool integrations from an open ecosystem.
Recall: Persistent Redis Memory for Claude (v1.5)
171
A Redis‑backed MCP server that gives Claude persistent, secure, cross‑session memory with powerful organization, search, and governance features.
After the GenAI Bubble: Fewer Layoffs, Persistent Hallucinations, and Pragmatic Code Gen
GenAI’s hype will pop: hallucinations persist, mass layoffs won’t happen, code-gen becomes a practical tool, and after the bubble bursts we’ll avoid the grifters’ future.
Designing Safe, Effective Agentic Loops for Coding Work
284
Safely empower coding agents to iterate autonomously by sandboxing YOLO mode, exposing simple shell tools, tightly scoping credentials, and relying on tests to guide trial-and-error.
Comprehension Debt: The Hidden Cost of Fast AI Code
532
Rapidly shipping unread LLM-generated code creates a mounting comprehension debt that will slow teams down when real changes are needed.
Claude Code: Terminal-Native AI Coding Agent with Easy Install and Privacy Safeguards
842
A terminal-native coding agent that accelerates development via natural language, easy to install and backed by clear privacy safeguards.
Claude Sonnet 4.5 Launches: SOTA Coding & Agent Model With SDK and Major Product Upgrades
1585
Anthropic unveils Claude Sonnet 4.5—its state-of-the-art, most aligned coding and agent model—alongside major product upgrades and a new Agent SDK, available now at the same price.
Avoiding the AI Coding Trap: Treat LLMs Like Fast Juniors with Real Engineering Discipline
685
Use AI’s speed within disciplined engineering practices—treat LLMs like fast juniors—to ship sustainable, high-quality software instead of quick but brittle code.
Coding Agents Don’t Lack IQ—They Lack Context
196
The bottleneck for autonomous coding isn’t IQ—it’s missing, implicit context that agents must access, synthesize, and query humans about.
Rebuilding a Startup Site with Claude: Fast, Powerful—But Human-Guided
178
AI can help non-engineers ship real, high-fidelity code fast—so long as humans stay in the loop to guide, review, and correct.
How HubSpot Scaled AI Coding: Context, Central Teams, and Data-Driven Rollout
Treat AI coding as a platform capability: measure it, centralize enablement, hardwire context, remove friction—and adoption will safely scale to unlock agents and bigger wins.
Zed Switches to Token-Based AI Billing, Cuts Pro to $10, Adds GPT‑5 and Gemini
182
Zed switches to token-based AI billing, cuts Pro to $10 with credits, adds top models, and offers flexible BYO/local options with a staged migration.
Make AI Work in Big Repos: Spec-First Workflow and Frequent Intentional Compaction
517
Make AI work in big, messy repos by compacting context and reviewing specs, not just code: research → plan → implement, with humans focused upstream.
Faster LLMs, Bigger Demands: Why Coding Agents Won’t Stabilize Soon
137
Faster LLMs will reshape coding workflows and productivity, but escalating demand, hardware limits, and pricing pressures mean a bumpy, fast-changing road ahead.
AI Amplifies Seniors, Not Juniors
461
Today, AI amplifies senior engineers’ impact instead of democratizing coding for juniors.
AI Agents Can Already Do Lean Proofs—But They Still Need a Project Manager
219
A general-purpose AI coding agent can already do real Lean proof engineering with guidance, hinting that theorem proving may soon be cheap and automated despite today’s rough edges.
Small, Business-Value Units Make AI Coding Work
170
Make AI coding reliable by breaking work into small, business-valued, human-verifiable units and rigorously engineering the context for each.
Microsoft steers VS Code to Claude 4, signaling an Anthropic tilt
213
Microsoft is steering VS Code and parts of Microsoft 365 toward Anthropic’s Claude where it performs best, even as it builds its own models and keeps working with OpenAI.
GPT‑5-Codex Lands in Codex Tools, API Coming Soon, With Big Push on Code Review
OpenAI’s GPT‑5-Codex is a tooling-first, code-focused upgrade that boosts review and refactoring while the API and polish catch up.
When Code Gets Cheap: Value Shifts to Judgment and Systems
115
As code gets cheap, the scarce—and valuable—skills become judgment, integration, and systems thinking, not typing more code.
GPT-5-Codex: Agentic Coding with Layered Safety
250
A safety-focused addendum introduces GPT-5-Codex, an agentic coding model trained on real tasks, widely available, and protected by layered mitigations.
LLMs Don’t Code—They Compile Your Prompts
410
LLMs don’t write code—they compile your prompts; treat them as tools and fix our languages and tooling instead of buying the hype.
Async Programming as a Workflow: Specify, Automate, Review
123
Define problems clearly, automate verification, and review thoroughly so AI can build in the background while you focus on higher-leverage engineering work.
Reviving a 1990s Linux Tape Driver with AI in Two Evenings
929
With careful guidance, an AI coding agent helped revive a 1990s Linux tape driver to run on modern kernels, proving AI as a strong force multiplier for legacy code.
AI Gatekeeper Slashes E2E CI Time by 84%
105
Let Claude Code act as an AI gatekeeper that inspects your PR and runs only the relevant E2E tests—cutting CI time by ~84% without losing coverage.
Disciplined AI Collaboration: Plan, Measure, and Ship in Small, Reliable Modules
Constrain AI with small, testable modules and continuous measurement to turn planning into reliable, data-driven delivery.
Today’s Tally: 16 “Absolutely Right” + 5 “Right”
651
A lighthearted dashboard counts how often Claude Code says he’s right—16 times "absolutely right" today plus 5 times "right."
Ship Faster by Orchestrating Parallel AI Coding Agents
Run many AI coding agents in parallel, orchestrate and review their work, and you’ll ship more by trading precision for throughput.
Ship Faster by Treating AI as a Forgetful Junior Dev
550
Use AI as a forgetful junior dev: provide rich context, expect three iterations, and enforce rigorous review to ship faster with better focus.
Senior Devs Ship More AI Code, Feel Faster—But Real Gains Are Mixed
215
Senior devs ship more AI code and feel faster, but real productivity gains are uneven and often offset by rework, even as enjoyment rises and sustainability concerns grow.
Vibe Coding, Not Replacement: A 40-Hour Test of AI as a Powerful but Unforgiving Pair Programmer
181
AI coding assistants dramatically accelerate development but demand expert oversight—vibe coding is a collaboration, not a replacement.