# Agentic Coding Weekly

> Get focused updates on agentic coding tools, models, and workflows. No hype, no gossip. Just what helps you ship.

This file provides information about Agentic Coding Weekly to help large language models understand and reference this publication's content.

## Posts
- [Talking to My Terminal with Local Speech-to-text and Pi Coding Agent](https://www.agenticcodingweekly.com/p/talking-to-my-terminal-with-local-speech-to-text-and-pi-coding-agent): Use local speech-to-text with Pi coding agent to turn spoken requests into shell commands and answers in your terminal.
- [ACW #24: Claude Fable 5 Access Pulled By US Government](https://www.agenticcodingweekly.com/p/acw-24-claude-fable-5-access-pulled-by-us-government): Also: Kimi K2.7 Code, GLM-5.2, MiMo V2.5 Pro UltraSpeed, and MiMo code.
- [ACW #23: MiniMax M3, Nemotron 3 Ultra, and GitHub Copilot Desktop](https://www.agenticcodingweekly.com/p/acw-23-minimax-m3-nemotron-3-ultra-and-github-copilot-desktop): Also: Qwen 3.7 Plus, usage-based GitHub Copilot billing, Open Code Review, and and a TDD skill worth bookmarking.
- [ACW Monthly Brief: May 2026](https://www.agenticcodingweekly.com/p/acw-monthly-brief-may-2026): Everything important in agentic coding this month, distilled for engineers and technical leaders.
- [ACW #22: Gemini 3.5 Flash, Antigravity 2.0, and Composer 2.5](https://www.agenticcodingweekly.com/p/acw-22-gemini-3-5-flash-antigravity-2-0-and-composer-2-5): Also: Antigravity CLI, Qwen 3.7 Max, DeepSeek V4 Pro's permanent discount, and aliases to stay sane across coding agents.
- [ACW #21: Claude Code Gets /goal and Agent View and Codex on Mobile](https://www.agenticcodingweekly.com/p/acw-21-claude-code-gets-goal-and-agent-view-and-codex-on-mobile): Also: Claude plan changes including +50% weekly limits, Grok Build CLI, and why you might want worse models on purpose.
- [OpenCode CLI Cheat Sheet](https://www.agenticcodingweekly.com/p/opencode-cli-cheat-sheet): A concise OpenCode CLI reference for agents, MCP, skills, config, workflows, and commands.
- [Claude Code CLI Cheat Sheet](https://www.agenticcodingweekly.com/p/claude-code-cli-cheat-sheet): A concise Claude Code CLI reference for commands, flags, MCP, skills, agents, hooks, and workflows.
- [ACW #20: Higher Claude Code Limits, Codex Goal Loop, and DeepSeek V4 Deals](https://www.agenticcodingweekly.com/p/acw-20-higher-claude-code-limits-codex-goal-loop-and-deepseek-v4-deals): Also: Simon Willison on the guilt of not reviewing AI code, running DeepSeek V4 Flash locally on Apple Silicon, and HTML for agents.
- [Codex CLI Cheat Sheet](https://www.agenticcodingweekly.com/p/codex-cli-cheat-sheet): A concise Codex CLI reference for commands, flags, MCP, skills, subagents, config, and workflows.
- [ACW #19: GPT‐5.5, DeepSeek V4, and Kimi K2.6](https://www.agenticcodingweekly.com/p/acw-19-gpt-5-5-deepseek-v4-and-kimi-k2-6): Also: Zed parallel agents, Claude Code postmortem, Copilot plan changes, and Roo Code shutdown.
- [ACW #18: GPT-5.4 Mini and Nano and Cursor Composer 2](https://www.agenticcodingweekly.com/p/acw-18-gpt-5-4-mini-and-nano-and-cursor-composer-2): Also: A practical multi-agent workflow and three reads on specs, codegen, and real-world usage.
- [5 Best Open-Source Speech-to-Text Tools in 2026](https://www.agenticcodingweekly.com/p/5-best-open-source-speech-to-text-tools-in-2026): Are you still typing every prompt? Use voice typing with Claude Code, Codex, Cursor, Copilot, or any coding agent to give richer context with less effort.
- [ACW #17: 1M Context for Claude and 2x Usage Promo](https://www.agenticcodingweekly.com/p/acw-17-1m-context-for-claude-and-2x-usage-promo): Also: Gemini CLI plan mode, JetBrains Air, `/btw` and voice mode in Claude Code, and 8 levels of agentic engineering.
- [ACW #16: GPT-5.4's 1M Context and Claude Code Scheduled Tasks](https://www.agenticcodingweekly.com/p/acw-16-gpt-5-4-s-1m-context-and-claude-code-scheduled-tasks): Also: Gemini 3.1 Flash-Lite, voice mode in Claude Code, agentic engineering patterns, and the debate over AI-assisted relicensing.
- [ACW #15: Claude Code Remote Control, Cursor Agents VMs, and Cloudflare Rebuilt Next.js](https://www.agenticcodingweekly.com/p/acw-15-claude-code-remote-control-cursor-agents-vms-and-cloudflare-rebuilt-next-js): Also: RTK CLI proxy to minimize LLM token consumption, two new skills in Claude Code, and voice dictation in Codex CLI.
- [Agentic Coding Weekly Quiz #4: Claude Code Subagents](https://www.agenticcodingweekly.com/p/agentic-coding-weekly-quiz-4-claude-code-subagents): Test your knowledge of Claude Code's subagent system and custom subagent configuration.
- [ACW #14: Claude Sonnet 4.6, Gemini 3.1 Pro, and Spec-Driven Development](https://www.agenticcodingweekly.com/p/acw-14-claude-sonnet-4-6-gemini-3-1-pro-and-spec-driven-development): Also: Taalas's LLM ASIC chip, building side projects with AI, and Claude Code subagents quiz.
- [Agentic Coding Weekly Quiz #3: CLAUDE.md](https://www.agenticcodingweekly.com/p/agentic-coding-weekly-quiz-3-claude-md): Five questions on Claude Code's memory system: hierarchy, rules scoping, imports, and --add-dir.
- [ACW #13: GLM‐5, MiniMax M2.5, and GPT‐5.3‐Codex‐Spark](https://www.agenticcodingweekly.com/p/acw-13-glm-5-minimax-m2-5-and-gpt-5-3-codex-spark): Also: Karpathy's 243-line GPT, optimizing agent harnesses, and a quiz on managing Claude Code memory.
- [Agentic Coding Weekly Quiz #2: SWE-bench](https://www.agenticcodingweekly.com/p/agentic-coding-weekly-quiz-2-swe-bench): SWE-bench Verified is the most most commonly reported coding benchmark. Test your understanding with 5 quick questions.
- [ACW #12: Claude Opus 4.6, GPT-5.3-Codex, and Agent Teams](https://www.agenticcodingweekly.com/p/acw-12-claude-opus-4-6-gpt-5-3-codex-and-agent-teams): Also: Qwen3-Coder-Next for local dev, Codex app for macOS, and building a C compiler with 16 parallel Claudes.