Ai-Coding

Claude Code officially ships native AGENTS.md support as of release 2.1.277 (September 18, 2026), automatically reading instructions from an AGENTS.md file when no CLAUDE.md is on the path. The feature arrives roughly 13 months after the most-upvoted issue in the repository was filed, and closes a saga that once saw the Shopify CEO publicly threaten to block the tool. The Timeline: From Most-Upvoted Issue to CEO Threat to Native Support The fight for AGENTS.md support in Claude Code is one of the defining public stories of the 2026 AI coding landscape. It started quietly on August 21, 2025, when a user filed issue #6235 asking Anthropic to make Claude Code read the cross-vendor instruction file. Over the following year, that single request became the most-upvoted issue in the entire anthropics/claude-code repository, accumulating roughly 4,667 to 5,026 thumbs-up reactions and 361 to 384 comments. ...

Claude Code weekly limits are higher in 2026 than they were at the start of the year, but the number that matters shifted on September 14: the 50% promotional boost that ran from May 13 to September 13 has expired, and Anthropic replaced it with a permanent 25% raise to standard weekly limits on Pro, Max, Team, and seat-based Enterprise plans. Relative to the promotional peak you may have grown used to, that is a drop of about 17% — so heavy users are essentially giving back part of their summer allowance while still keeping more than their original baseline. This guide walks through the full 2026 timeline, the exact math, and practical tactics to stretch the limits you actually have. ...

Coding agents like Claude Code and Codex remember nothing once a session ends, which is why you keep re-asking the same questions and re-discovering the same fixes. Lore is a local-first desktop app that reads the session files these agents already save, connects them to your repositories and Git history, and puts everything into a single searchable SQLite archive — with no accounts, no telemetry, no cloud database, and no LLM calls. Because it only reads what Claude Code and Codex already write on disk, Lore adds searchable long-term memory to your coding agents without changing how they behave or locking you into another vendor. ...

Can you really build a custom smart watch face with an AI coding agent on a $27 watch? Yes — a staff software engineer at Strava did exactly that, turning a PineTime that had sat abandoned in a drawer for two years into a working Casio-style watch face in just a few hours. This case study breaks down how edge AI coding works on cheap, open-source hardware, and what it means for the future of embedded development. ...

You can absolutely use AI to code at home without going broke. By combining free-tier cloud tools like GitHub Copilot Free with open-source local models running on Ollama and Continue.dev, indie developers can build a fully functional AI coding assistant for $0 per month. A one-time investment of $200–500 in a used GPU or an Apple Silicon Mac unlocks local models that match cloud services on most everyday coding tasks, saving you hundreds of dollars per year compared to subscriptions like Cursor Pro or Claude Team. ...

Kimi K3, Moonshot AI’s 2.8-trillion-parameter Mixture-of-Experts model, has rapidly become one of the most talked-about coding models of 2026. In hands-on testing, K3 optimized a Rust function using SWAR (SIMD Within a Register) in about 15 minutes with zero errors, achieving a 2% program-wide speedup at a cost of roughly $1 on OpenRouter. With open weights scheduled for release by July 27, 2026, and pricing at roughly one-third the cost of Claude’s top-tier models, K3 presents a compelling case as the new value leader in AI-assisted coding. ...

On June 17, 2026, JetBrains moved Junie out of beta and shipped three additions that change the calculus for anyone deciding between Junie, Cursor, and Claude Code: agentic debugging with native IDE breakpoint control, a standalone CLI for CI/CD pipelines, and bring-your-own-model keys. I’ve been running the GA release for three weeks across IntelliJ IDEA and PyCharm, and the agentic debugging feature is the one that makes Junie genuinely different from every other AI coding agent on the market right now. ...

Trae AI IDE is a serious Cursor alternative if price is your main constraint, but I would not treat it as a default team IDE yet. In 2026, its $10 Pro plan, SOLO mode, MCP support, and VS Code base are compelling. Its model restrictions and telemetry story need a harder look. What Is Trae AI IDE? Trae AI IDE is ByteDance’s AI-first development environment built on the VS Code ecosystem. That matters because the migration story is familiar: editor layout, extensions, terminal workflows, keyboard habits, and many Code OSS assumptions carry over better than they would in a completely new IDE. ...

The best Claude Code dev team stack in 2026 is small: project CLAUDE.md rules, 6-8 focused skills, 4-5 read-first MCP servers, deterministic hooks, and subagents for noisy investigation. I’ve found that teams get more value from governing those layers than from installing every popular skill. That matters because Claude Code has stopped being “just a terminal chatbot.” The current extension surface includes skills, hooks, subagents, agent teams, code intelligence plugins, MCP servers, and packageable plugins. The temptation is to install a marketplace worth of tools on day one. In practice, that usually makes the system slower, harder to trust, and harder to debug. ...

Claude Code subagents are the cleanest way to delegate noisy, self-contained coding work without filling your main session with logs, search results, and half-finished reasoning. In 2026, they are also the simplest entry point to parallel AI development: Markdown files, YAML frontmatter, scoped tools, optional worktree isolation. I use subagents when a task has a clear boundary and a noisy execution path. Running a full test suite, researching three unrelated modules, auditing a diff for security issues, or generating migration notes are good fits. Asking a subagent to co-own an ambiguous refactor with you is usually a bad fit. The difference matters because subagents do not share your full conversation by default. They start with their own prompt, their own context window, and a task summary from the main agent. ...