A collection of 32 posts
Qwen3.8-27B posts benchmark wins over Claude Opus 4.6 Max and ships a first-party Claude Code launcher via Ollama. We grade the claims, fix the overthinking latency problem, and redraw the hybrid local-vs-cloud line.
Cohere North Mini Code 1.0 is an open-weight 30B MoE coding model (3B active, 256K context, Apache 2.0) built for agentic software engineering. Specs, benchmarks, access.
Qwen3.7-Max is the closed flagship with higher vendor benchmarks and a 1M context; Kimi K2.7 Code is the open-weights, cheaper agentic specialist. We compare access, benchmarks, pricing, local feasibility, and which to use for autonomous coding.
Claude Code adds abstractions, fallbacks and files nobody asked for. Here are the practical controls - CLAUDE.md rules, plan mode, scoped prompts, effort settings and review gates - that get you surgical changes instead.
OpenCode just passed Claude Code on GitHub stars. We compare the open-source, model-agnostic terminal agent against Anthropic's polished CLI on setup, models, MCP, pricing, and who should pick which.
Run Claude Code against OpenRouter instead of Anthropic billing: the three-env-var setup, which open models work as backends, live pricing, and the tool-use and prompt-caching catches the hype threads skip.
A realistic, no-hype guide to running Qwen 3.6 27B locally as a Claude Code alternative: benchmarks vs Opus, the hardware and quant you actually need, how to wire it in, where it holds up, and what a hybrid setup actually saves you.
OpenAI's Codex CLI quietly logged at TRACE level to a local SQLite database, writing an extrapolated ~640 TB a year to some developers' SSDs. Here's what happened, how to check whether your drive was affected, and how to fix it.
CLAUDE.md is Claude Code's per-project memory file, auto-loaded into context at the start of every session. Here is how to write one that actually helps: anatomy, an annotated example, common mistakes, and how it relates to AGENTS.md and Cursor rules.
Zhipu Z.ai shipped GLM 5.2 today on every GLM Coding Plan tier with a usable 1M-token context window. Standalone API, the Z.ai chatbot, and the MIT open weights are arriving next week. No benchmarks yet — here's what's confirmed, what's not, and how it fits next to GLM-5.1.
Running five Claude Code sessions in five terminals and forgetting which one's stuck waiting on you? A free board that gives every agent its own column, a task queue, and a notify-only signal when one needs your input.
How senior engineers wire Claude Skills and MCP servers together in 2026: SKILL.md format, the MCP 2025-11-25 spec, real integration patterns for code review, database access, and incident response.
How Cursor Composer, Claude Code, Codex CLI, and Gemini CLI compare on setup, agents, MCP, models, and pricing in 2026.
AGENTS.md, CLAUDE.md, .cursor/rules, SKILL.md, and Copilot instructions all do the same job differently. Here is the 2026 breakdown of format, frontmatter, monorepo handling, and which one to pick.
In a single month, xAI shipped a coding agent, a skills system, and a connectors layer. Here's how Grok Build 0.1, Grok Skills, and Platform Connectors fit together — and how the stack compares to Claude Code, Cursor, and Copilot Workspaces.
Quick answer. Gemini CLI is Google’s open-source terminal AI agent. Install it with npm install -g @google/gemini-cli on Node 20+, run gemini, and sign in with your Google account in the browser. The free tier covers 60 requests per minute and 1,000 per day with Gemini 2.
Xiaomi's open-weight MiMo-V2.5 and the 1.02T-param V2.5-Pro target agentic coding. Here is an honest, sourced read on whether they match Claude.
Use the Playwright MCP server with Claude Code and Cursor to generate E2E tests against your live DOM — setup, review loop, failure modes, conventions file.
Kimi K2.6 vs DeepSeek V4 vs GLM-5.1 for coding in 2026 — sourced benchmarks, real cost-per-task, self-host and license comparison, plus a clear pick-X-if decision block.
A structured checklist for vetting developers who code with AI in 2026 — judgement, code-review-of-AI-output, and agent fluency over leetcode.
A senior-engineer walkthrough of running Cline or Continue.dev fully local against Ollama on a 24GB GPU — exact config, VRAM tables, and an honest quality reality check vs cloud Claude.
Install the Cline agent in Cursor and point it at DeepSeek V4 — real config JSON, Pro vs Flash guidance, the reasoning_content gotcha, and a cost worked example.
Grok Build, Claude Code, and Codex CLI compared on benchmarks, pricing, and workflow — including Anthropic's June 15 metered-credit change.
Install and run Grok Build, xAI's new terminal coding agent: real install command, auth, Plan Mode, models, pricing, and how it stacks up.
Anthropic's June 15, 2026 billing change moves Claude Agent SDK and claude -p onto a separate metered credit. Here's the exact pre-deadline checklist.