# Give to Claude: Run the first Multica request

You are Claude Code working with Multica after the four MVP agents and skills have been created.

Your job is to create/run the first useful parent request through the Chief of Staff and verify that the team loop starts correctly.

## Safety

- Do not publish anything publicly.
- Do not edit a production website.
- Do not spend money.
- Do not claim rankings, traffic, or commercial outcomes without evidence.
- Any publish/send/deploy/customer-facing action must be marked `Approval needed`.

## Step 1 — Create the parent issue

Create one parent issue assigned to Chief of Staff.

Use this request, replacing the example domain if the human supplied a real one:

```text
Do an SEO audit on https://example.com and have the team take actions to help us rank better.

Give me back a clear report on what you found, what the team did, and what needs approval before anything goes live.
```

## Step 2 — Expected Chief of Staff behaviour

The Chief of Staff should:

1. Own the parent coordination issue.
2. Create relevant child tickets.
3. Assign SEO Consultant to audit the site.
4. Assign Researcher only if evidence, competitor context, source material, or buyer language is needed.
5. Assign Blog Writer only if the audit supports content/blog opportunities.
6. Require readable HTML reports/previews where useful.
7. Return a final package with actions, evidence, artifacts, gaps, and approvals.

Do not manually choreograph all child tickets unless the Chief of Staff fails to route them. The point is to test the role/skill setup.

## Step 3 — Monitor the loop

Check that:

- parent issue exists
- child tickets are created where useful
- specialists receive clear briefs
- no agent tries to do everything
- reports are readable and evidence-labelled
- no public/published/production action happens without approval

## Done when

- The first request is created and assigned to Chief of Staff.
- The Chief of Staff has started or completed the routing loop.
- Any blockers are reported clearly with command/action, error, and next human step.