Home › Free Marketing Starter Kit
Free and open source · For Claude Code

The PELORA Marketing Starter Kit for Claude Code

A free, ready-made folder that turns Claude Code into a small marketing team: a rules file, four role files, six skills, a reviewer that reads your work as your customer would, two safety hooks, and a morning report you can schedule. Download it, fill in your facts, and start.

  • Free
  • Open source
  • MIT license
  • For Claude Code

MIT license. No email needed for the kit. Not affiliated with or endorsed by Anthropic.

The ideaWhat is the PELORA Marketing Starter Kit?

It is the setup from our guide, Claude Code for marketing: how to build the system, packed into one folder. It is built from how PELORA Marketing runs its own work with Claude Code, with everything private taken out. Claude reads the rules at the start of every session, works inside one role at a time, checks its drafts with a reviewer, and asks a person before connector calls that look like spending or publishing.

What you getWhat is inside the kit?

PartWhat it does
CLAUDE.mdThe rules Claude reads first, every session: who you are, who you sell to, how work moves, what it must never do.
brand/One home for your facts, voice and banned words. Claude quotes facts from here, never from memory.
seats/One file per job (we call them seats): ads, SEO, email and reporting.
Six skillsResearch first, ad copy, SEO page, email campaign, weekly report and creative brief. Claude loads one only when the task needs it.
ReviewerA subagent that reads each draft as your customer and returns READY or NOT READY with the exact fix.
Two hooksA claims check that blocks a draft in your content folder when it uses a banned phrase, and a spend guard that goes by each connector tool's name: anything not named as a simple read, like get_ or list_, waits for a person's yes (a budget, bid, launch or send, for example). Both need Python 3.
Morning reportA prompt for a cloud scheduled task: what came in overnight, and what the platforms changed. Cloud scheduled tasks work from a GitHub repository. Run the prompt once by hand first, so no read is left waiting for a yes.
healthcare/A health claims list and an ad review checklist for clinics and health brands.

Your first afternoonHow do I start with the kit?

  1. Install Claude Code from Anthropic's live docs (code.claude.com/docs) and sign in. Most ways to use it need a paid Claude plan or an Anthropic Console account.
  2. Check Python 3 is installed: type python3 --version in a terminal. If it is not found, install Python 3 first. The two hooks need it.
  3. Unzip the download and open the folder in Claude Code: in the desktop app's Code tab, or in a terminal type cd and the folder's path, then claude.
  4. Fill in your facts and voice in brand/. Leave blanks where you cannot prove something yet.
  5. Test the hooks: ask Claude to write "This is guaranteed" into content/test.md. The claims check should refuse. Delete the test file. Then ask Claude to run python3 .claude/hooks/test-hooks.py: both lines should say OK.
  6. Run a first task: "You are in the ADS seat. Write three search ads for our main service." Then type: "Use the customer-eye-reviewer on that file."
  7. Each week, add one line to CLAUDE.md for each mistake you never want repeated.

Health businessesCan a clinic or health brand use the kit?

Yes, for marketing work: ads, pages, email to people who opted in, and reports built from platform numbers. Keep patient information out of Claude Code unless your organization has a business associate agreement with Anthropic, runs a setup that agreement covers, and every tool or connector it uses is covered too. Anthropic's Claude Code legal page lists which setups qualify. Turn on the health claims list in the kit, and keep your own compliance review. The list blocks a draft that uses a listed phrase so it gets rewritten; it does not make anything compliant.

Want the step-by-step walkthrough?

The PDF walkthrough takes you through five short sessions, about an hour in all: your facts, the safety test, a first draft, one connected account, and the morning report. No email needed.

Download the walkthrough (PDF)

Rather not build itCan PELORA set it up for us?

Yes. We build the full system inside your business, across ads, search, email and reporting, and teach someone on your team to run it. See how we build your marketing in-house with AI.

Start with a call

Book a free 15-minute strategy call. We look at how your marketing runs today and tell you what we would set up first.

Book a Strategy Call

QuestionsFrequently asked questions

Is the kit really free?

Yes. It is open source under the MIT license. The kit and the walkthrough both download with no email.

Do I need to know how to code?

No. You fill in plain text files and type requests in plain English. The two hooks are short Python scripts, so Python 3 has to be installed. You do not change how they work; you only list your connector's write tools in one of them, and the README shows where.

Does the kit work with Google Ads and Meta?

Yes, through Claude Code's connectors, with your own sign-in. Connect with read access where the platform allows it. The spend guard goes by each tool's name, not by what the tool does. Any connector tool whose name is not a simple read, like get_ or list_, has to wait for your yes. After you connect an account, ask Claude: "List every tool this connector has and mark any that can change something." Add those names to EXTRA at the top of .claude/hooks/spend-guard.py, or set STRICT = True there so every connector call asks. The guard does not watch scripts or direct API calls, so do not keep a write-access API key in the folder.

Is this an official Anthropic product?

No. PELORA Marketing made it. We are not affiliated with or endorsed by Anthropic. Claude and Claude Code are trademarks of Anthropic.

Last updated November 16, 2026. By Preston Durnford. Newport Beach, California.