My Top Picks
Not sponsored. These are the tools I'd genuinely recommend to anyone building with AI agents — the ones I reach for every day and would point a colleague to without hesitation.
Essentials
Terminal-based, works with any project. I've built 19 specialized agents on top of it for a full autonomous development pipeline — from Jira ticket to merged PR.
This is where agentic engineering actually happens. Not just autocomplete — full workflow automation, multi-agent orchestration, persistent memory across sessions.
My agent pipelines push code; Actions validates it. Runs tests, checks formatting, deploys on green. Simple, reliable, integrates with everything.
Essential infrastructure for agentic workflows. An agent that can't get feedback from CI is flying blind.
Worth Adding
A GitHub Actions workflow that calls the Claude API on every pull request. Claude reads the diff, understands the context, and posts a review comment before a human looks at it.
One of the highest-leverage additions to a team's CI pipeline. Catches issues early, keeps quality high without slowing down review cycles.
Run LLMs locally — Llama 3, Mistral, Qwen and more. Zero API cost, full privacy. Useful for high-volume, low-stakes automation where cloud quality isn't required.
I use it for tasks like generating AI comments at scale. Not the first tool you need, but a useful addition once you're running real workflows.