Kousigan A

Series

5 parts

AI-Assisted Engineering

Working with AI coding agents, one part per post: lay out the project, plan before code, pick skills, agents or MCP, verify the output, write reusable skills.

Start with part 1

  1. part 1Deep dive10 min

    Beyond AGENTS.md: the six layers of a coding agent project

    How to structure an AI coding agent project beyond one instructions file: rules, skills, sub-agents, MCP, project docs and checks, plus three prompts.

  2. part 2Deep dive9 min

    Don't let your coding agent code first: plan, then approve

    A coding agent workflow with a stop before any code: investigate, plan, approve, implement and verify, plus the plan modes of seven agents.

  3. part 3Deep dive11 min

    Skills vs agents vs MCP: don't build an AI maintenance job

    Skills, sub-agents, MCP and hooks solve different problems. When to use each in a coding agent, where six agents keep them, and when to add nothing.

  4. part 4Deep dive11 min

    "Looks good" isn't evidence: how to verify AI-generated code

    AI-generated code is not done when it compiles. Check the requirement, the diff, tests and static checks, and get a review from outside the model that wrote it.

  5. part 5Deep dive9 min

    Not every repeated prompt deserves an agent skill

    How to tell which repeated coding agent work deserves a skill, what a good SKILL.md holds, and how to keep it from going stale, with two prompts to copy.

All topics and tags