Two days to make AI part of how your team ships
We build software with AI agents every day. It is how we move faster and deliver more while staying inside a project's budget. In two days, on site or remote, we check the basics with your developers and then set up the workflows that make the difference, on your own repositories. The course is tool agnostic: we support Claude Code, Codex, Copilot, Cursor and Gemini CLI, and we tailor the content to your team, your needs and your tools.
Solutions
We scope every engagement to your situation. This is what we usually deliver.
Workflow and tool choice
Where AI coding tools stand today, IDE assistants against terminal agents, which model for which task, and what each one costs.
Context and instruction files
AGENTS.md and IDE rules as the team's living documentation. Project and personal memory, path-scoped rules, and how to diagnose an agent that drifts.
Automation and guardrails
Custom commands, skills and hooks: reusable prompt recipes, automatic formatting, blocked dangerous operations, and the checks that run before a commit.
Integration and delegation
MCP servers for your tools, subagents for review, tests and documentation, planning before code, and headless runs in CI.
The two days
A typical split. We adjust it to your team's level and tools after a short call.
Day 1: foundations
Where the tools stand
A short history, the IDE versus CLI split, and how to pick a tool and a model for each task.
Your environment
Install and configure the tool your team uses, with its instruction files at the root of your repository.
Context that works
Diagnose context problems, write memory the agent follows, and keep it from growing stale.
Safe experimentation
Checkpoints and rewinds, permission modes, and a plan reviewed before any code is written.
Spec first
From a ticket to a specification, a plan and an implementation your team reviews.
Day 2: automation on your repositories
Commands and skills
Turn the team's repetitive prompts into commands and skills everyone shares.
Hooks
Formatting, linting, tests and blocked operations that run on their own, on every change.
MCP
Connect the agent to your tools: issue tracker, database, browser and design files.
Subagents
Delegate review, tests and documentation to specialized agents, and keep the cost in check.
CI and headless runs
Automated review on pull requests, pipelines that fix themselves, and the agent inside your CI.
Book the course for your team
Tell us about your team, your stack and your tools. We come back with a program and a date.
Request the course