| name | rhombus-partner-onboarding |
| description | Onboarding runbook for adding a new client organization to partner management. Covers provisioning API access, inviting the partner admin, setting retention policy, configuring alert routing, and validating the onboarding with a smoke test. Use whenever the user says "add a new client", "onboard an org", "new MSP client", "set up a new managed org", "client X is going live", or asks to walk through the steps for bringing a client under partner management.
|
| argument-hint | [client-name] |
| allowed-tools | Read, Bash, Write |
Partner Client Onboarding
Runbook for bringing a new client organization under partner management. Every step is a CLI call; the checklist lives in references/onboarding-checklist.md.
Prerequisites
- You are authenticated as a partner (
rhombus partner get-partner-clients-v2 returns your client list).
- The client organization already exists in Rhombus (i.e., the client has spun up their own Rhombus org via the Rhombus Console, or Rhombus sales has provisioned it).
- You have the client admin's email ready (for the partner-admin invite).
High-level flow
- Confirm the client org is visible —
rhombus partner get-partner-clients-v2 | grep -i "<client-name>". If not present, ask the user to reach out to Rhombus to attach the client to their partner account.
- Set as active client —
/rhombus-client-switch <client-name>. Every subsequent command scopes to this client by default.
- Provision the partner API key for the client — this is what the partner uses to automate against the client. The
rhombus-cli auto-generates one on rhombus login; for scripted usage, use the Developer Webservice.
- Invite the partner admin to the client org with the role that matches your engagement (usually "Admin" for full MSP, "Viewer" for monitoring-only).
- Set retention and alert policies — bring them in line with your MSP's standard baseline. Ship the baseline as a saved JSON at
references/onboarding-checklist.md so every client starts identical.
- Configure alert routing — webhooks to your monitoring platform (PagerDuty, Slack, custom), email lists, and escalation rules.
- Validate with smoke tests — pull recent alerts, fetch camera list, simulate a test alert, confirm it lands in your monitoring.
- Document the onboarding in your client-ops log.
Walkthrough
Parse $ARGUMENTS as the client name. If missing, ask.
Then for each step in references/onboarding-checklist.md:
- Announce the step.
- Run the command.
- Show the output, flag any red flags.
- Ask the user if they want to continue, make a change, or abort.
End with a summary report of every action taken, ready to paste into the client-ops log.
When to prefer the slash command
/rhombus-clients to discover the client, /rhombus-client-switch to activate, then invoke this skill by asking Claude to "onboard ". The rhombus-client-onboarding agent wraps this into a guided flow.
Edge cases
- Client already has an existing partner integration. Don't clobber. Ask the user how to handle it (override, side-by-side, abort).
- Retention policy conflicts with a client's existing configuration. Always surface the diff before applying — never silently overwrite retention settings.
- Webhook URL not yet provisioned. Note as a gap; complete the rest of onboarding so you can deliver a working-baseline-minus-one-thing.