Agentic coding guides for developers
Practical, tool-agnostic guides on working with AI coding agents. They apply whether you use Claude Code, Codex, or Cursor. New to the terms? Start with the glossary.
- Context engineering for coding agents: what to load and what to guard
Most agent tokens are old context re-sent on every call. Learn how a coding agent uses its context window and how to keep it small and relevant.
- How to review agent-written code without piling up tech debt
Agent-written code needs a different review. What to check once the tests pass, which patterns create tech debt, and how to keep changes small.
- Subagent orchestration: plan with a strong model, execute with cheaper ones
Orchestration lets a stronger model plan and verify while cheaper models do the narrow work. How to set it up and the mistake that makes subagents expensive.
- What is agentic coding? A practical guide for developers
Agentic coding means an AI agent plans, edits, runs tools, and checks its work while you set the goal and verify the result. Here is the system behind it.
- Validation loops: how to make agent-written code verifiable
A validation loop is an automated check the agent runs on its own changes. Learn which checks to use, how to wire them in, and how to enforce them.
- How to write an AGENTS.md that coding agents actually follow
Keep AGENTS.md short, specific, and enforced. What to include, what to leave out, and why instructions alone are only suggestions to an agent.