Claude Cowork, Spec-Driven Development, and Ralph Wiggum
And more AI Coding news: January 14, 2026.
Hey hey,
This is Povilas Korop from AI Coding Daily.
Another very busy week in AI Coding world, with a lot of news and tutorials.
So, I decided to double-down on curating them and launched a new official website: AICodingDaily.com!
On that website, you will find AI coding tutorials, prompts, tools, and digest of latest posts from X / Reddit.
For now, I will publish content both here on Substack and for website audience, but essentially, I grew out of Substack and want more structure and control over how content is presented. I will give more information about that website next week, including perks for Premium members.
Now, let’s dive into this week’s content!
From my AICodingDaily.com website
[TUTORIAL] Disable Claude Code Commit/PR Attribution (Clean Git History)
If you ask Claude Code to commit and push changes to GitHub in a prompt, it will add itself as a co-author of that commit. How to remove it?
[NEWS] Claude Code Restricts Subscription Access to OpenCode and Others
On January 9th, people using Claude Code subscription inside of OpenCode, started seeing a message “This credential is only authorized for use with Claude Code”.
My YouTube videos
NEW Claude Code Plugin: Code Simplifier (4 min)
Boris from Claude Code announced a new plugin, and I decided to try it out.
Claude Code Prompt: Review AI Agent Work Against Specification (4 min)
Another great prompt by Thariq from Claude Code team.
Spec-Driven Development for AI Agents: I Tried OpenSpec and Others (28 min)
A longer video, showing OpenSpec in action, discussing other tools, and showing our own prompts-driven approach for Laravel projects.
I Tried Ralph Wiggum Plugin for Claude Code (9 min)
Let me show you the approach of Ralph Wiggum methodology, applied as a Claude Code plugin.
The Future (I think) of CLAUDE.md and Other Guidelines (8 min)
I think I realized something pretty important about AI rules and guidelines. They are not (that much) anymore about the code.
How I Use Claude Code Web on My Phone (For Small Changes) (4 min)
Let me show you how I make small changes to my project when I have ideas while away from computer.
From AI Coding Community
x.com
Introducing Cowork: Claude Code for the rest of your work.
Cowork lets you complete non-technical tasks much like how developers use Claude Code. In a separate tweet, Boris from Claude Code said that Claude Code generated ALL of Cowork’s code, and they built it in 1.5 weeks.
burkeholland.github.io
Opus 4.5 is going to change everything
Three months ago I would have dismissed claims that AI could replace developers. Today, after using Claude Opus 4.5, I believe AI coding agents can absolutely replace developers.
x.com
Introducing Claude Delegator
A Claude Code plugin that lets you use GPT 5.2 powered subagents directly within Claude Code.
x.com
In opencode v1.1.11 you can now use your ChatGPT Plus/Pro plans
This is kind of an answer to Claude Code restrictions for Opencode, which I talked about in the news article above.
x.com
Claude Code 2.1.0 is officially out!
Shift+enter for newlines, w/ zero setup, Add hooks directly to agents & skills frontmatter, and much more
x.com
Claude Code v2.1.1 introduces a new “Bash” subagent
A command execution specialist. Useful for avoiding context pollution in the main thread by moving multi-step operations to another context window.
x.com
More updates dropping with Claude Code 2.1.3!
Merged slash commands and skills - does this mean they’ll share the same features?
x.com
Cursor: “Here’s what we’ve learned from building and using coding agents.”
This guide covers techniques for working with Cursor’s agent. Whether you’re new to agentic coding or looking to learn how our team uses Cursor, we’ll cover the best practices for coding with agents.
x.com
Tip by Ado: “Building an app that imports smth outside the current directory?”
Claude Code is sandboxed to your current directory. What to do?
x.com
The Complete Guide to Building Agents with the Claude Agent SDK
The Claude Agent SDK is engine, yours to point at whatever problem you want, so you can easily build agents of your own.
youtube.com
My Claude Code Workflow for 2026
20-min video by Ray Amjad.
x.com
Tip by Jarrod Watts: “This rule prevents Claude from writing slop comments.”
It prevents Claude from writing slop comments like “increment counter” on already self-documenting code.
x.com
Cursor’s agent now uses dynamic context for all models
It’s more intelligent about how context is filled while maintaining the same quality. This reduces total tokens by 46.9% when using multiple MCP servers.
youtube.com
I Found a Better Way to Use Claude Code (Long Running Agents)
28-minute video from Leon van Zyl.
youtube.com
Claude Code Plan Mode Plugin - Plannotator
2-min video. Interactive plan review, sharing, and automated feedback for Claude Code.
That’s a wrap for this week. Keep building with AI!
Povilas Korop
AICodingDaily.com



This roundup is gold, especially the spec-driven development stuff for Laravel projects. The idea of using OpenSpec alongside your own prompts-based approach is facinating becuase it forces you to think thru project architecture before letting AI loose on the codebase. I've been wrestling with how to structure my Laravel apps when Claude Code keeps wanting to jump stright to implementation. Congrats on the new website launch btw, been following your content for a while now!