The 60-second version

A ClawCurrent kit is a folder of instructions, templates, and checklists for a specific business — for example a roofing company or a dental practice. It is not software. It is the operating manual that an AI agent (like OpenClaw, running on your computer or server) can read and follow.

After you buy, you hand the kit to your agent. The agent reads it, builds a workspace on your machine, and starts producing real work: draft replies to new leads, a content calendar, follow-up emails, weekly reports. Everything sensitive — sending, publishing, spending, connecting accounts — waits for your explicit approval first.

You are not buying a chatbot. You are buying a documented way of working, plus an agent that executes it, with you as the manager who signs off.

What happens after checkout, step by step

1

You download the kit

It appears in your account instantly. Inside: a README, an agent manifest (agent-product.json), install guides (AGENTS.md, SKILL.md, INSTALL.md), workflows, templates, and QA checklists.

2

Your agent reads the kit

You point your harness (OpenClaw, Agent Zero, MemUbot, or any AGENTS.md-compatible runner) at the folder. The agent reads the manifest and tells you what inputs it needs — business name, service area, prices, tone.

3

The agent builds your workspace

It creates organized folders for intake, drafts, workflows, QA, and reports. Nothing is public yet — everything is a draft on your machine.

4

The agent starts producing drafts

Within the first session it can produce things like: a response template for new quote requests, a 4-week content calendar, a follow-up email sequence, and a launch QA report.

5

You approve; then it runs on a rhythm

Once you approve the first outputs, the kit's recurring workflows kick in — weekly reports, pipeline summaries, content drafts — always drafted, never sent, until you say so.

The agent roles inside a kit

Think of these as job descriptions, not separate products. One agent can play several roles; the kit tells it what each role does and when.

Business Installer

Setup

Reads the kit, creates your workspace folders, and asks you for the details it can't guess. Example: "What's your service area and hourly rate?"

Lead Intake Agent

Revenue

Turns new inquiries — form fills, calls, quote requests — into drafted replies for you to approve. Example: a same-day response draft to a homeowner asking for a roof quote.

Content Pipeline Agent

Growth

Builds content calendars, service-page briefs, and email campaigns around your business model. Example: four weeks of local SEO post ideas, drafted for review.

Trust & QA Agent

Safety

Checks claims, links, and compliance risks before anything customer-facing leaves the workspace. Example: flags a review quote you don't have permission to use.

Reporting Agent

Operations

Turns activity into weekly summaries: what was drafted, what's waiting on you, what's next. Example: a Monday-morning report of open loops and recommended actions.

Connection Guide Agent

Setup

Walks you through connecting accounts (website, CRM, calendar, email) with least-privilege access — and documents exactly what each permission allows.

Where your approval is always required

Agents draft and prepare. The following actions never happen without your explicit yes:

Requires approval — Granting or changing account permissions

Requires approval — Connecting accounts or credentials

Requires approval — Publishing pages, posts, ads, or listings

Requires approval — Sending customer-facing email, SMS, or chat

Requires approval — Spending money or changing budgets

Requires approval — Legal, medical, financial, or compliance-sensitive claims

Ready to run

Pick the business. The kit handles the rest.

Every kit includes the manifest, install guides, workflows, QA checklists, and approval gates described above — plus a 14-day money-back guarantee. New to all of this? Start with the plain-English explainer.