Engineering workflows for Claude Code and AI coding agents
A tested pack of skills, agents, native command files, prompts, and checklists for engineers who want agentic coding to be repeatable, reviewable, and less chaotic.
Planned launch price: $39. The public claude-skills repo remains free and useful.
Inspect before editing
Turn a vague repo into a file-level map, risk list, commands, and smallest safe next action.
Fix with evidence
Keep debugging tied to the failing command, hypotheses, relevant files, and verification.
Review before shipping
Find behavioral regressions, missing tests, data risks, and unclear rollback paths before merge.
Standardize team usage
Give engineers shared prompts, command files, and checklists instead of one-off agent habits.
What is inside
The pack is not a bag of prompts. It is a working operating layer for engineering agents: command files, skill folders, agent roles, validation notes, and proof examples.
- 20 curated engineering skills
- 5 pack-specific agents
- Native Claude Code command files
- Repo inspection, focused fix, review, TDD, and ship-gate workflows
- Prompt sets, checklists, install notes, and validation scripts
- Before/after proof examples from real repositories
Get the first release
Join the launch list for the Engineering Agent Skills Pack. You will get the free starter path first, then the paid pack when the final proof pass is done.