A blank scaffold to start your team's shared workflow layer. Companion to the Claude Code for Orgs workshop.
The same folder structure as the Faculty Products Org OS example — but empty. Drop your team's content into the right folders and you've got a working Org OS.
.
├── CLAUDE.md # the lean root — fill in your team
├── customers/ # CONTEXT: customer calls, summaries, playbooks
├── analytics/ # CONTEXT: metric definitions, queries, playbooks
├── templates/ # CONTEXT: PRD, status update, RFC structures
└── .claude/
├── skills/ # ACTIONS: shared workflows (slash commands)
└── rules/ # BEHAVIOR: file naming, prioritization, etc.
gh repo create my-org-os --template carlvellotti/org-os-starter --private
cd my-org-osOpen CLAUDE.md. Replace the placeholders with your team:
- Team roster (names, Slack handles, GitHub handles)
- Slack channels you actually use
- Folder pointers (these probably don't change)
Keep it under one page. The rule is lean root, nested files for everything.
Pick one piece of context that's currently in your head or in old Slack threads. Write it down in the appropriate folder:
- A customer? →
customers/account-summaries/ - A metric definition? →
analytics/metrics/ - A playbook? → wherever it fits
Just one. The OS grows from real friction, not pre-planning.
Pick the most-duplicated task on your team. Write it as a skill in .claude/skills/. Use the example skills in the FP repo as a reference.
This is the whole game. See the /orgos page for the full rollout playbook.
CLAUDE.md— root template with placeholder team infocustomers/CLAUDE.md— folder navigation map templateanalytics/CLAUDE.md— folder navigation map templatetemplates/— empty.claude/skills/— empty.claude/rules/— emptyCONTRIBUTING.md— example contribution guide for your team to adapt
- /orgos workshop page — the full walkthrough
- Faculty Products example repo — populated worked example
- The Full Stack PM newsletter — weekly on AI + product
Built by Carl Vellotti.