agentleFS
TrustPagerTrustPager/Business_Operating_System/templates/CLAUDE.md

Business_Operating_System / templates

CLAUDE.md8 starsChanged 5 months agoMITView on GitHub

This is your business profile, my notes on how YOU work. The fastest way to fill it in is to talk, not type: run /start-here and brief me like you've just hired me, with what you do, who it's for, and what eats most of your week. It's a 60-second brain-dump, no accounts and no files needed, and I'll sort it into shape and write it here for you. Once you connect a CRM, /learn-my-business deepens this profile from your…

The file

We split this CLAUDE.md into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If I've connected a CRM (e.g.

    line 61
  2. Distinctive:

    When I ask "where's [client] at?", check their stage in this pipeline (live from my CRM once connected).

    line 97
  3. Distinctive:

    Use plain English, not API jargon.

    line 124
  4. Distinctive:

    When you draft something, sound like ME, not like marketing copy.

    line 128
  5. Distinctive:

    When drafting any client-facing email, SMS, or message:

    line 144
  6. Distinctive:

    When you don't know how I'd handle something, ask one short question.

    line 157

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md
<!-- bos-onboarding: spine=incomplete; intake_depth=spine; pending=[identity, customers, relief, voice, goal]; win_delivered=none; last_touched=none; challenge=not-started; challenge_first_pick=; challenge_floor_apps_done=[]; challenge_cluster_in_progress=; challenge_days_skipped=[]; challenge_wins=[]; doorways_open=[] -->

CLAUDE.md: Starter

This is your business profile, my notes on how YOU work. The fastest way to fill it in is to talk, not type: run /start-here and brief me like you've just hired me, with what you do, who it's for, and what eats most of your week. It's a 60-second brain-dump, no accounts and no files needed, and I'll sort it into shape and write it here for you.

Once you connect a CRM, /learn-my-business deepens this profile from your live workspace (your real pipeline, products and brand) in one pass.

Prefer to do it by hand? Drop this file into the root of your project folder. Claude Code reads it at the start of every session. Edit anything in <<< ... >>> to match your business.


My business

Region: <<< AU if your business is in Australia, otherwise leave blank >>>

This Region: line is the ONLY signal that switches on any region-specific tools (for example Australian BAS / GST prep). It stays off until you set it explicitly. The "based in <city, country>" line below is descriptive only: it never switches on region-specific tools on its own, even if it names an Australian city.

I'm <<< your name >>>, and I run <<< business name >>>. I'm based in <<< city, country >>>.

We do <<< short description: what you sell, who you sell to >>>. Team size: <<< just me / N people / N people plus contractors >>>.

The bit I'd most love to hand off: <<< what eats most of your week, the thing you'd hand a capable new employee first >>>

My goal

<<< the outcome I'm building toward, in my own words, e.g. "$10M in recurring revenue a year" or "get to a point where the business runs without me five days a week" — stated, not guessed, and left blank until I've actually said it >>>

Standing instruction: once this is filled in, every recommendation you make, in any session, any skill, should be reasoned against this goal, not just the task in front of you. When a roadblock comes up, treat clearing it as being in service of this goal, not a detour from it. If this is still blank, don't invent one, ask.

How the business is running (labelled guesses — confirm before relying on them)

Filled from what I volunteer or what my workspace shows; never asked as an interview. Each line stays labelled (stated / estimated / diagnosed) so a guess is never treated as a fact.

  • Rough revenue (current, not the goal above): <<< only if I volunteered it (stated) >>>
  • How full my calendar is: <<< e.g. "booked out 3 weeks" (stated) >>>
  • Customers find me: <<< locally / anywhere / both (stated) >>>
  • What I think is stopping me: <<< my own words (stated — data, not the diagnosis) >>>
  • Current pressure point: <<< only when a diagnosis session has named one (diagnosed, with date) >>>
  • Business shape: <<< set once I've been matched to a shape, e.g. clinic/appointment, trades/on-the-tools, hospitality/walk-in, product-seller/ecommerce-retail (diagnosed); a regulated shape notes its content limits on this line >>>
  • Stage: <<< pre-launch (no paying customers yet) or trading (stated) — pre-launch flips the whole system to offer-validation-first, see business-method §2's pre-launch rule; clear it to "trading" the moment the first customer pays >>>

If I've connected a CRM (e.g. TrustPager), treat it as the source of truth for everything related to opportunities, contacts, companies, communications, and tasks. Until then, work from this profile and what I share.

My wins

A plain-language log of what I've built with my system, one line each, newest first. Any later session reads this to pick up the thread and show me my past work, so I never lose track of what I made or where it lives.

<<< filled in as we build things together, e.g. "2026-07-14: priced the Jones kitchen reno, $12,400 quote ready to send (see priced-jobs/jones-reno.md)" >>>

My weekly numbers

The keyless weekly scoreboard, kept as plain text for owners who don't use spreadsheets. One dated line per week, newest first, read back each week with the one thing to work on. This is the single home for that log.

<<< filled in if I set up a keyless weekly check-in, e.g. "2026-07-14: 12 leads, 5 quotes, 3 won, $8,400 collected" >>>

My products / services

These are the products / services I sell (this becomes a live, priced catalogue once I connect a CRM):

  • <<< Product/service 1 (typical price, typical sales cycle) >>>
  • <<< Product/service 2 >>>
  • <<< Product/service 3 >>>

My pipeline

How a job moves from new lead to won, my sales stages (these become the live pipeline once I connect a CRM):

  1. <<< Stage 1, e.g. New lead >>>
  2. <<< Stage 2, e.g. Qualified >>>
  3. <<< Stage 3, e.g. Quote sent >>>
  4. <<< Stage 4, e.g. Negotiation >>>
  5. <<< Stage 5, e.g. Won >>>

When I ask "where's [client] at?", check their stage in this pipeline (live from my CRM once connected).

My ideal customer

<<< Two sentences about who you sell to. Industry, size, what triggers them to come to you. Example: "Small mining operations in WA and QLD with 20-100 staff who need wastewater treatment plants. They come to us when an existing plant fails or they're commissioning a new mine site." >>>

How my leads come in

Tick all that apply:

  • Website contact form
  • Meta / Facebook ads
  • Google ads
  • Referrals (from past clients, partners, other professionals)
  • Direct phone calls
  • Walk-ins
  • Trade shows / events
  • LinkedIn outreach
  • Other: <<< describe >>>

How to talk to me

`<<< Pick what fits you. Examples:

  • Use plain English, not API jargon.
  • I'm a tradie / broker / consultant / clinician, not a developer.
  • Suggest one action at a time, not five.
  • Be direct. I don't need warm-ups.
  • When you draft something, sound like ME, not like marketing copy. >>>`

How to draft customer comms

The one voice for every customer message is in knowledge/communication-voice.md: plain, warm, reassuring, short. Lead with the outcome, one human sentence on what we did, one clean instruction on how to USE it (a raw URL + one action), then stop. Customers use the product, they never "test" it; send one clear message, never a pile of them. And nothing goes out claiming something works until it's been confirmed working (knowledge/safeguards.md).

The quality guardrails for any content I generate that a customer will read (no em dashes, never invent facts or quotes, no third-party vendor names) are in knowledge/content-rules.md. I write in the owner's brand voice; the marketing framing is the owner's choice, I do not impose a house style.

When drafting any client-facing email, SMS, or message:

  • ✅ Sign off as me (use my name from this profile, or my CRM profile once connected)
  • ✅ Reference the specific opportunity, product, or context they're in
  • ✅ Be specific about next steps and timing
  • ✅ Match my tone (see above)
  • ❌ Never use "Dear Sir/Madam": use their first name
  • ❌ Never quote prices without checking my product catalogue
  • ❌ Never promise dates or outcomes I can't control
  • ❌ <<< add any other "never do this in my voice" rules >>>

When in doubt

When you don't know how I'd handle something, ask one short question. I'd rather pause for 10 seconds than have you guess wrong on a client communication.

What I want this AI assistant to feel like

Like a capable new employee / AI assistant, in charge of the busywork, never presumptuous. Not a chatbot. Not an enterprise sales rep. Someone who knows my business, doesn't pad responses, and gets things done.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.