How are AI coding agents evolving through 2027? A practitioner breaks down the trends, an L0-L5 autonomy ladder, and where kits fit for Claude Code.
Blog · Jasmine Daily
Everyday moments, told honestly .
Where I write down thoughts, experiences, and the little kind things picked up along the way - not rushed, not perfect, just a personal journal.
Showing 10 posts.
Showing 10 posts
What is context engineering in AI coding? Learn context windows, token budgets, and the 4 strategies to curate context so AI writes better code.
AI slop is AI code that looks right but quietly rots your codebase. Learn the warning signs and a checklist to keep quality while vibe coding.
Learn how to debug with AI: use Claude Code to trace the real root cause instead of patching symptoms, with a real session, prompts, and 6 clear steps.
AI code review catches bugs on the diff before you merge: how to enable review, see real finding examples, and cut false positives fast.
How to do TDD with AI: write tests first with Claude Code using red-green-refactor, with a real test-first session and how to force the AI to comply.
Spec-driven development means writing a spec and plan before AI writes code. Get a real spec.md and plan.md, a 6-step workflow, and run it in Claude Code.
The vibe coding workflow from idea to ship with Claude Code: brainstorm, write a spec, plan, cook, test, and deploy. Includes a real example. Start now!
What is vibe coding? Learn how AI writes code from plain English in 2026: the workflow, tools, pros and cons, and when not to use it.
Turn Claude Code into a repeatable AI dev workflow: brainstorm, plan, cook, ship with AgentKit commands and real screenshots. See the full workflow.
Jasmine Daily
Thanks for stopping by.
If it resonates, come back often - I write every week.