Tier 2 · Team · $15K-$35K · 2-4 weeks
A two-to-four-week engagement that gets five to twenty-five people productive with Claude Code on shared conventions. Team-wide CLAUDE.md standards, a shared skills library, MCP servers everyone uses, and 30 days of async support.
The Problem
This is what we see most often. One person on the team starts using Claude Code. They get good at it. Word spreads. Three months later, half the team has it installed, and every person has a slightly different CLAUDE.md, a slightly different set of skills, and a slightly different opinion on what works.
That is not a rollout. That is drift. And it is the thing that eventually kills adoption, because no one can share a setup, no one can help the next hire, and the same skill gets re-invented in four slightly broken ways.
The Team Rollout gives your team one shared setup, built together, that everyone actually uses.
The Engagement
Week 1: Team audit and kickoff
We interview five to seven people across your team. Who already uses Claude Code. How. What works. What does not. What tasks eat the most time. This gives us the evidence we need to pick the right shared patterns instead of imposing ones that do not fit.
Week 1-2: Shared CLAUDE.md and settings standards
Team-wide instructions everyone agrees on. Project-level and user-level. Settings.json templates distributed across the team. Permission allowlists and denylists that match your security requirements. Hooks for the things everyone wants to happen automatically.
Week 2-3: Shared skills library
Six to ten skills built in a shared repo. These are the recurring tasks across your team: code reviews, PR summaries, ticket drafts, meeting notes, client reports. Anyone on the team can invoke them. Anyone can improve them. Version-controlled.
Week 2-3: Shared MCP servers
The data sources your team shares: GitHub, Linear, your internal wiki, your ticketing system, your codebase search. Set up once, used by everyone. Where useful, we build a custom MCP server for an internal API so Claude has context on your actual systems.
Week 3-4: Training sessions and handover
Three live sessions. Kickoff (all staff, 90 minutes), mid-review (working session, 90 minutes), and optimisation (final tune-up, 60 minutes). Plus a handover doc your team owns: what is installed, how to extend it, and how to onboard the next hire.
What You Get
A Git repo with your team's skills, versioned and documented. When someone improves a skill, everyone benefits. When you hire someone, they clone the repo and they are productive on day one.
One template everyone starts from. With project-specific overrides. So the tone, conventions, and guardrails are consistent across the team without flattening individual style.
The shared data sources connected and working. Documentation so a new hire can reproduce the setup in thirty minutes.
A simple dashboard showing who is using the skills, how often, and which ones are earning their keep. Gives you evidence instead of vibes when you report adoption to your boss.
Recorded. Your team can rewatch them. Your next hire can watch them during onboarding. They become the institutional memory of how your team decided to use Claude Code.
A shared Slack channel (or your equivalent). Questions get answered within a business day. Skills get tuned. New ones get scoped. After 30 days, you have the option to roll into a retainer.
Who Books This
Engineering teams are the most common client. Five to twenty-five developers, using Claude Code for code review, PR writing, test generation, and internal tooling. The team rollout standardises patterns across the squad.
We also run this for operations teams, marketing teams, and consulting practices where the staff are doing knowledge work and want to share their best AI workflows rather than each reinvent them. The skills are different (client reports, research briefs, proposal drafts) but the rollout shape is the same.
If your team is larger than 25 or you need SSO, custom MCP servers for multiple internal systems, and API-scale workloads, step up to the Implementation Partnership. If you are rolling out to one person first, the Power User Day is the better start.
How Pricing Works
Team size is the main driver. A team of five engineers with a simple codebase is toward the lower end. A team of twenty across three functions, needing a custom MCP server for a bespoke internal system, is toward the upper end.
We quote fixed fee after a 30-minute scoping call. You know the price before you commit. No hourly billing that mysteriously expands.
For teams under five people, the Power User Day per person is usually more cost-effective. For teams over twenty-five, the Implementation Partnership is the right step.
No more parallel setups. No more drift. One shared library that makes every next hire faster to onboard.