#Development #Techniques
Using AI to write better code more slowly · “Try this other, slower style of ‘vibe coding.’” https://ilo.im/16d7sk
_____
#Programming #Coding #Debugging #CodeReviews #Workflows #WebDev #Frontend #Backend
#Development #Techniques
Using AI to write better code more slowly · “Try this other, slower style of ‘vibe coding.’” https://ilo.im/16d7sk
_____
#Programming #Coding #Debugging #CodeReviews #Workflows #WebDev #Frontend #Backend
This is going to gain me some unfollows, or at the very least some side-eye/eye-roll looks.
The more I use #LLM tools in my day to day, the more I like them. Not to #VibeCode, but to use them as an actual #assistant.
I already wrote a couple of #automations for my team to improve our workflows and improve #consistency and #quality. One we are all actively using to interact with Jira, by following our team standards when writing epics, stories, spikes etc., as I said, it helps keep the content consistent and aligned.
I am also working on an automation for me that could eventually become another team-wide one that gives me a high-level status of all my pull requests, pending #CodeReviews, issue status, etc. and gives me a brief paragraph with how my day is going and recommends what should be my focus (based on the given priorities).
Do I need LLMs for this? No, but as a good engineer, the less I have to worry about some things, the more bandwidth I have to worry about other important things.
Could I do something like this with standard scripting? Most of it yes, but there is a very nice result of using LLMs that would be extremely hard or time consuming with standard tools; they are especially good at adapting to small edge cases and unforeseen or unimplemented features.
I also have to say I am not a stranger to vibe coding. I've done it, and I really like it for those little personal scripts and one-offs that used to take so long with so little gains.
Anyway, if you want to unfollow, block or just eye-roll me, I won't hold it against you.
AI Code Reviews
Discover how adamsreview simplifies multi-agent PR reviews for Claude Code
Fast by Default, by @denodell:
AI Amplifies Everything: A Team Lead’s Guide to AI-Assisted Development, by @frontendmasters.com:
AI-Assisted Coding: A Practical Guide for Software Engineers, by @frontendmasters.com:
https://frontendmasters.com/blog/ai-assisted-coding-a-practical-guide-for-software-engineers/
Discover how #Meta improved #SoftwareQuality with a Just-in-Time (JiT) testing approach that dynamically generates tests during code review.
The system increases bug detection by approx 4x in AI-assisted development using LLMs, mutation testing, and intent-aware workflows like Dodgy Diff.
More on #InfoQ ⇨ https://bit.ly/4tqylKc
#SoftwareArchitecture #SoftwareTesting #AI #LLMs #CodeReviews
#Anthropic has introduced a new Code Review feature for Claude Code, adding an agent-based pull request review system that analyzes code changes using multiple AI reviewers.
Dive deeper on #InfoQ ⇨ https://bit.ly/3QbwdHA
GitHub Introduces Stacked Prs to Ease Review Bottlenecks, by @techstronggroup:
https://devops.com/github-introduces-stacked-prs-to-ease-review-bottlenecks/
There’s a Hidden Tax on Every AI-Generated Merge Request, by @brianwald.bsky.social (@TheNewStack):