One page. Short enough to remember, clear enough to settle an argument.
A one-page expense policy for a small team, built from the questions people actually ask and tested on five real expenses.
Teams without a policy renegotiate every expense, and the answer depends on who is asked. Teams with a long policy have one nobody opened since it was written. Both end up with questions in chat and quiet resentment about what counts.
Expense Policy starts from the five to eight things your team really asks about. It sets one limit and one approver for each, says what receipt is needed and where it goes, and ends with a single exceptions line. Then it runs five real expenses through the page to see where it fails.
What it produces.
- →A one-page policy under about 350 words
- →A category table with a limit and approver for each
- →Receipt, submission, and reimbursement rules in three lines
- →A five-case test showing where the draft is still unclear
- ✦Owners of small teams with no written policy
- ✦Office managers fielding the same expense questions weekly
- ✦Growing businesses appointing their first approvers
- →expense-policy.zip: a native Claude Skill. Upload to Settings → Skills in Claude, or unzip to
~/.claude/skills/ - →expense-policy-skill.md: plain markdown. Works in ChatGPT Custom GPTs, Gemini, Claude Projects, agent system prompts, anywhere
- →A README inside the zip with install steps for Claude, ChatGPT, and Gemini
What it does with a real question.
We're a 12-person design studio and we've never had a written expense policy. Last month our ops manager, Priya, had to chase down receipts for $4,300 in client dinners and software subscriptions. I'm the founder and currently approve everything. Starting in March, two team leads will approve too. Can you draft a one-page policy?
Start with the two things that cost you $4,300. Priya's chase was about client dinners and software subscriptions, so those two categories get the tightest rules, and the page stays short enough that nobody skips it. Everything else can wait for a later revision.
For client dinners, I'd set a cap per head, say $90, require a named client, and make you the approver every time, since you want that spending visible. For software, a new subscription needs a team lead's yes before purchase, while renewals of tools you already use need nothing. That one rule would have prevented most of Priya's pile, because a subscription on a personal card with no prior approval is exactly the kind that arrives later without a receipt.
Pick one pre-approval line for everything else, such as $200, with team leads approving under it and you above it. Say in one sentence that receipts get photographed and uploaded within ten days, and reimbursement happens in the next payroll run after approval. Only promise a timing Priya can actually keep.
Finish with a single exceptions line: ask a team lead, and if you cannot reach one, spend the minimum and tell them the same day.
Before March, run five of last month's actual expenses through the draft. Any case that still needs someone to ask a question shows you which sentence to fix.
The full Playbook runs the whole procedure and writes the complete deliverable, not this excerpt.
Written by claude-sonnet-5-5 with this Playbook loaded, 2026-10-11. Checked automatically for length and for copied framework text.
No subscription · free updates within v1.x
Works as a Claude Skill (.zip) and as a plain .md for ChatGPT, Gemini & every LLM
Delivered to your email within 60 seconds · Secured by Stripe
Price includes UK VAT where applicable
Before you buy.
How is this different from just asking Claude directly?
Asked cold, Claude writes a long, formal policy that covers every case, which is the kind nobody reads. This keeps it to one page built from your actual disputes, uses your own limits rather than invented ones, and tests the draft on real expenses before you publish it.
How is this different from a Tool?
Playbooks are systematically produced to a fixed template and priced accordingly ($12.99). Tools are hand-built, one at a time, and go deeper ($39 to $199). A Playbook gives you a reliable method for routine work; a Tool is for when the judgment matters more than the format.
Will this work with ChatGPT, Gemini, and other LLMs?
Yes. You get a native Claude Skill (.zip) that auto-invokes in Claude, plus a plain .md you can paste into a ChatGPT Custom GPT, a Gemini system prompt, or any LLM. The procedure travels with the file.
Can I share this with my team?
It's licensed for your own personal and professional use. The output it produces is yours to circulate. To put the Playbook itself in the hands of a whole team, email hello@authority.md about team licensing.
What if it doesn't work for me?
If the file hasn't been delivered, or arrives defective or empty, we refund in full: see our refund policy.
Finance Playbooks are writing and structuring aids. They are not financial, tax, or accounting advice. Check figures, policies, and anything with legal or tax consequences with a qualified professional.