CritterWatch tenant operations Added in 1.14.0
Operating tenant lifecycle on a CritterWatch-monitored service through the MCP tenant tools — add_tenant, enable_tenant, disable_tenant, remove_tenant and hard_delete_tenant — with the caution each one deserves, and list_tenants to look first. Covers what each tool really does on the service (fan-out across every registered IDynamicTenantSource, CREATE DATABASE on add, DROP DATABASE ... WITH (FORCE) on hard delete), why accepted:true means published and not applied, the confirmation protocol for destructive calls, the traps (hard delete of a DISABLED tenant is refused — silently before CritterWatch 1.1; remove_tenant destroys data on a partitioned conjoined source; hard delete is PostgreSQL-only; one hard delete drops every source's database), the capability split between tenant.remove and tenant.hard-delete, and the audit row the MCP path lacked before 1.1.
Use when an agent is asked to onboard, suspend, offboard or permanently delete a tenant, or to explain what one of these tools will do before it is called.
Skill details
- Skill name:
wolverine-integrations-critterwatch-tenant-operations - Applies to:
wolverine,marten,polecat - Tags:
critterwatch,mcp,multi-tenancy,tenants,operations,destructive-actions,rbac,agents,marten,efcore - Added in: 1.14.0
- License: Proprietary. Copyright JasperFx Software LLC.