Your AI coding assistant is smart but inconsistent: it forgets your conventions. Skills Pack for Builders gives it 14 job descriptions — SKILL.md files in the Agent Skills standard, auto-loaded per task: smart-commit, code-review, debug-detective, scaffold-feature, test-plan, refactor-plan, docs-writer, idea-validator, pricing-guide, gumroad-listing, landing-page, launch-checklist, build-in-public, changelog-writer. 30-second install; MIT licensed. Launch deal: $19 for 3 days, then $29.
Hey PH! I'm Ab, a software engineer with 5 years of experience.
I built this because I was tired of re-explaining the same conventions to my AI assistant in every session. Write tests first. Commit messages like this. Checklist before you say "done." The model knows how, it just never remembers what "done" means on my team.
So I wrote each expectation down once, as a skill, and now the agent loads the right one on its own. The code-review skill alone has caught things I would have shipped.
Everything is MIT licensed, installs in 30 seconds, and works with Claude Code, Cursor, Copilot, Codex, and VS Code. PH launch deal: it is $19 for the next 24 hours, then back to $29. Happy to answer anything about the Agent Skills standard or how I use these day to day.