Remco Brilstra’s Post

GitHub is bringing remote control to Copilot CLI, letting developers run and manage coding sessions beyond the local terminal. The move mirrors what Anthropic already introduced with Claude Code, as both push toward agents that can run and be steered from just about anywhere. 🔗 Read the full story here: https://lnkd.in/gfP2iQQ7

I'm all-in with Claude Code to write most of my code. Using it to review that same code always felt wrong. Yes, GitHub Copilot has a review feature & Claude Code can comment on a PR too via their App. I've also set up the Cursor CLI inside a GHA workflow to make reviews with different models. But those are all tools doing a reviewer's job on the side and I didn't manage to get really good results, regardless of the efforts on the system prompt / instructions - probably a layer 8 problem, not gonna argue here. What's working for me finally... using a dedicated tool. CodeRabbit was built for exactly one thing and it does a MUCH better job that everything I've tried before. And now it lives in my terminal too. Their CLI can review your staged changes before you commit & not after you created the PR. The two things I like most: 1) It works with Claude Code directly. When it flags something, you hand it off to Claude to fix in one command. 2) Reviews happen locally before the "remote cycle" begins. If you prefer the IDE, the VS Code extension (also works in Cursor obviously) does the same thing (it's also free). Both are free and come with generous rate limits! https://fandf.co/48la8fA This post was created in collaboration with CodeRabbit, but it's a genuine recommendation after weeks of usage! 💛

Built a quick CI/CD cheatsheet while working through Github Actions on a personal project. 15 minutes to shift from "familiar with" to "can reason about". It's the kind of gap-filling that compounds when you're building full-stack applications.

Working with Claude Code for a while now—mainly via CLI at work and the claude.ai/code interface for personal projects. Picked up a few skills that turned out to be genuinely useful, sharing here: * action-run — Trigger and monitor any GitHub Actions workflow_dispatch by name, collect inputs, and auto-handle failures * action-check — Check latest GitHub Actions runs, monitor in-progress jobs, and surface failures with logs * check-pr — After pushing, track PR checks until completion, highlight unresolved comments, and report failures shortify — Clean up markdown docs (reduce verbosity, duplication, and noise) * git-commit — Create commits from staged changes, including pre-commit hook handling * local-skill — Pull a single skill directory from a GitHub repo into .claude/skills/ (great for claude.ai/code) to Install: /plugin marketplace add kolodkin/devpowers /plugin install devpowers@devpowers Feel free to try it out and share.

doqtor v0.1.1 is now available on npm. We built doqtor to solve the problem of documentation drift. Instead of generating docs from scratch, it monitors your code changes and ensures your existing documentation stays accurate. It identifies signature mismatches, renamed symbols, and outdated examples, then opens pull requests with the fixes automatically. Key features: - Automatic drift detection for TypeScript projects. - Surgical updates that only process what changed. - AI-powered fix generation for complex documentation updates. - Available as a CLI and a GitHub App. The project is still evolving and we are actively working on new features. We would love for you to try it out, share your feedback, or contribute to the codebase. Website: https://lnkd.in/gzRt3wwJ NPM: https://lnkd.in/gdiK7crc GitHub: https://lnkd.in/gqbrCNir

New video: a full beginner's guide to Claude Code. If you've heard about it but haven't actually sat down and used it, this is the one. What's covered: Setup. VS Code extension, authentication, opening the right folder. Plan Mode. Tell Claude to read your workspace and outline architecture before writing code. The setting that prevents most bad builds. Bypass permission mode. /config and check "dangerously skip permissions" so it stops asking approval on every command. The Chrome DevTools MCP. Once installed, Claude opens Chrome, screenshots the app it just built, identifies its own UI bugs, and fixes them. Context window management. When to /clear so you don't lose detail mid-build. Build along: a Prompt Vault app, dark mode, local storage, sort and save your prompts. https://lnkd.in/g2aEUsE9

Free-tier users really get no love 😅 Just ran a Plan + Start implement (4/5), and hit the limit right at the final validation step. With Agentic coding, a lot of the real learning honestly comes from paid usage. Without spending a bit, it’s hard to figure out the best way to actually use these tools. So far, GitHub Copilot still feels like the best option. But it really depends on your development workflow — you need to break down tasks well, plan things out, and feed enough context in one go to get good results.

Built and published a custom PTO and shift-management web app from the ground up using Python, Flask, SQLite, HTML, and CSS. The system includes PTO request handling, shift swap workflows, command-staff review tools, site-based access controls, audit history, notifications, calendar visibility, and user account management. I also deployed the live version to Pythonanywhere and created a separate public GitHub version of the project. This was a great project for strengthening full-stack development, deployment, workflow design, and real-world internal tool building. GitHub: https://lnkd.in/em6kGAZr

Mostly, I do a couple of things simultaneously: write code, do code reviews, and manage the team. Code reviews take a significant amount of time because I need to think through architecture decisions. Some things like code style, naming, and clean implementation can be partially enforced by linter rules, but not everything. So I decided to create a few skills to help review my code and my colleagues’ code. There are three of them: - Flutter Architecture Guard: boundary checks, dependency direction, and ownership clarity. - Flutter Design Pattern Review: practical pattern detection, anti-pattern analysis, and minimal structural recommendations. - Flutter Simplify: behavior-preserving refactors for readability and maintainability. These skills help me a lot and significantly reduce the time I spend on code reviews. Public repo with the skills: https://lnkd.in/e_rPT_Qk

Anthropic's Claude Code source code just got leaked 🚨 It’s now Public !! 23K+ stars on the leaked Claude Code source code GitHub Repository in just 5 hours. Crazy !! The world's most powerful coding agent was not able to secure its own code 🤷♂️😳🤯 Misconfigured .map file in their npm package exposed everything. #AI #Agents #LLM #RAG #code #software #web #development