ads-client-onboarding
Skill1 starsChanged 23 days agoMITView on GitHub
Add or update an advertising client step by step - business, niche, offer, economics, service area, brand colors, platforms and account ids, billing check, tracking, budget plan (daily vs monthly, how much), approval mode, autopilot on/off, scheduled reports. Use for "add <client> as a new ads client", "set up ads for my business", "onboard", or when adsgm resolve finds no matching client.
In this repository17 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- llms.txtllms.txt
- ads-auditSkill
- ads-budgetingSkill
- ads-communitySkill
- ads-competitor-researchSkill
- ads-coreSkill
- ads-creative-analysisSkill
- ads-landing-pageSkill
- ads-optimizationSkill
- ads-portfolioSkill
- ads-reportingSkill
- ads-strategySkill
- ads-trackingSkill
- google-adsSkill
- meta-adsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 30
Ask: per day, per month, a fixed promotion, or "you tell me"?
- Very specific: line 31
Run adsgm budget-plan --client <slug> ...
- Distinctive: line 65
Never paste tokens or card details into a profile.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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.
Files
Client onboarding
Goal: one validated profile at ~/.ads-godmode/clients/<slug>.json (outside the repo, never committed). Template: templates/client-profile.json.
How to run it
- One group per message, 2–4 questions. Reflect back, then wait for "yes, correct" before the next group.
- Anything done on a website gets a clickable link and the exact clicks. The links are in docs/SETUP_WIZARD.md, Steps 3–6.
- "Not sure" is a fine answer: record it empty and move on. Never guess an account ID.
- Progress line on every message:
Client setup · 3 of 8 · Money.
The groups
- Business: name, website, what you sell, the one offer to advertise first, the niche (short slug, e.g.
roofing-storm,dental-implants), service area (cities + radius), excluded areas. - Customer and economics: target customer; average job value; how many leads become customers (close rate); margin. If they don't know a target cost per lead, compute the most they can afford (job value × close rate × margin × ~30%) and show the math (HEURISTIC).
- Brand (for graphic ads): main, background and accent colors (hex, or "take them from the website"), logo file path, a folder of real job photos. Approved reviews they're happy to quote.
- Accounts:
- Meta ad account (
act_…), Page, Instagram, dataset/pixel. - Google customer ID (and manager ID).
- GA4, Search Console, CRM.
- Missing any? Walk them through creating it with the wizard links.
- Verify every ID with a read and confirm the account name with them.
- Meta ad account (
- Billing check: does each ad account have a payment method? If not, link them to Meta Billing & payments / Google Ads Billing. They enter the card themselves; you never see or type it. Offer to suggest an account spending limit after the budget plan.
- Money:
- Ask: per day, per month, a fixed promotion, or "you tell me"? Is the month a hard ceiling? Which platforms?
- Run
adsgm budget-plan --client <slug> ...and explain in plain words: budget type, what it buys, whether to start on one platform, and whether to spend a bit more (with the math) or hold back. - Agree on the plan, then
--saveit.
- Control:
- Approval mode:
approvalis recommended;draftif they only want builds. - Who approves, and the high-budget threshold.
- Banned claims and required disclaimers.
- Special ad category concerns (financing, housing).
- Approval mode:
- Autopilot and reports. Always offer both.
- Autopilot, on or off: it pauses clear losers, fixes and relaunches tired ads in the same ad set, and blocks a few wasted Google searches.
- It stays under a daily spend cap they choose (suggest the plan's daily figure plus 10–20%), and never launches new campaigns, raises budgets or touches billing.
- If yes:
adsgm autopilot on --client <slug> --cap <amount>. - Scheduled reports (daily check, Monday review, monthly report): if yes,
/ads:schedule.
Save
adsgm clients new <slug> # template in ~/.ads-godmode/clients/
# fill it (Write tool) with the answers
adsgm clients validate ~/.ads-godmode/clients/<slug>.json
adsgm clients check-isolation # no account may belong to two clients
adsgm use <slug>Then offer /ads:research <client> (niche playbook) and a first read-only /ads:audit.
Finish with a summary:
- client, niche and platforms;
- accounts, with the verified names;
- billing OK or not;
- the budget plan;
- the approval mode;
- autopilot and schedules, on or off;
- what's missing.
Never paste tokens or card details into a profile. Never copy one client's account IDs into another.
Several clients in one chat
Onboard each one in turn, as separate profiles. Afterwards the person can mix requests in one chat ("roofing campaign for A, dental campaign for B"). Run adsgm use before each client's work, and the gate checks every change against that client's accounts.
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
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.

