USE CASE / AI ENABLEMENT
Make AI enablement part of how your team works.
01 / BRAINS
Give AI coding agents the context your team already earned.
Brains bring team knowledge into a shared, version-controlled home. Preserve decisions, corrections, and solutions so engineers and their agents can retrieve relevant lessons for the next task.
For the enablement lead: organize the context a team needs, keep ownership clear, and give new engineers a useful starting point. This is context engineering grounded in your own work.
/BrainsEngineeringEngineering onboarding
Start with service boundaries and the shared API client.
Authentication and retries belong in one place.
Use the existing error format for new endpoints.
02 / SESSIONS
See where AI adoption helps and where engineers get stuck.
Sessions capture the work behind an outcome: goals, decisions, summaries, agent activity, and the skills used. Review what happened and preserve context for a handoff.
For the enablement lead: start with real friction. Use session evidence and token usage to choose the next workflow to improve, then pair those signals with code quality and engineer feedback.
/SessionsEngineeringHow the team uses its agents
Claude Code
Codex
Cursor03 / SKILLS
Turn a useful approach into a skill the team can reuse.
Share agent skills with version control and access control. Give a repeatable procedure a clear home so the team can maintain it as the workflow changes.
For the enablement lead: capture the steps behind a useful practice, choose who should use it, and keep the procedure current. Shared skills give training a practical follow-through in daily AI coding work.
/SkillsEngineeringBuild an onboarding brain
Turn repo history into a dated Brain for the person joining.
Claude CodeCodexCursor04 / RULES
Put reviewed standards into the agent’s workflow.
Review and approve proposed rules, then apply them in supported workflows to guide agent actions, request approval, or block actions.
For the enablement lead: make AI coding governance concrete. Choose the behavior that needs a check, test the rule, and review exceptions before expanding its use.
/RulesEngineeringCreate a team rule
Claude CodeManagerWhen a file has more than 300 lines, delegate the read to a cheaper model or read only the lines needed.
Delegate large file reads
Team rule · Needs review
Ready for the team’s next session
YOUR FIRST WORKFLOW
Start with one team and one repeated problem.
Use Sessions to understand a recurring correction. Preserve the explanation in a Brain. Document a repeatable procedure as a Skill. If a rule is appropriate, review and test it before applying it. Then compare the next set of similar tasks with your baseline.
Measure repeated corrections, review rework, and the effort the process adds. Developer productivity improves when the work improves; activity and token counts alone do not prove that.
Already have an AI enablement lead? Give them a foundation to build on. Still hiring? An existing engineering owner can start a focused pilot and preserve the decisions for the incoming lead.
AI enablement, explained.
What does an AI enablement lead do?
An AI enablement lead helps a team adopt AI tools, improve everyday workflows, share useful practices, and evaluate results. In engineering, that work can sit within Developer Experience, Developer Productivity, or Platform Engineering.
How do Brains, Skills, Sessions, and Rules work together?
Brains hold shared knowledge. Skills describe reusable procedures. Sessions provide evidence about the work. Rules apply reviewed standards in supported agent workflows. Together, they give an enablement lead a practical foundation for ongoing improvement.
Can we start before hiring an AI enablement lead?
Yes. Assign an existing engineering owner to one scoped workflow, document the baseline, and preserve the decisions for the incoming lead. The team still needs someone to own priorities, training, and rollout.