What tools can KAI call?

mcp.db.migrate: Apply database migrations. mcp.api.deploy: Deploy the API to staging/prod. mcp.queue.worker: Start the background worker. mcp.infra.terraform: Apply infrastructure as code. mcp.monitor.grafana: Set up monitoring dashboards

VZU OS · Agent

Agent 03 of 12

KAI.

Backend + infra engineer.

Kai is the VZU Stack agent. Full-stack engineering: database, API, queue, worker, frontend, infra, deploy, monitoring. The agent you call when the work is foundational.

Brief VZU on KAI → Open stack.vzu.ca →

The role

What KAI does

KAI is the dedicated agent behind VZU Stack. It runs every time the operator briefs the vzu stack capability. It owns the goal, the toolset, the memory, the audit trail.

The goal

What KAI is trying to achieve

"Stand up a multi-tenant, production-grade full-stack system with database, API, frontend, infra, monitoring, and observability."

The capability

What it powers

KAI powers the VZU Stack capability in the VZU family.

Visit stack.vzu.ca →

The toolset

5 MCP actions KAI can call.

Every action is exposed as an MCP server (Model Context Protocol). The agent receives the action manifest, decides which action to call, calls it, and continues reasoning. The audit trail records every call.

mcp.db.migrate

Apply database migrations
mcp.api.deploy

Deploy the API to staging/prod
mcp.queue.worker

Start the background worker
mcp.infra.terraform

Apply infrastructure as code
mcp.monitor.grafana

Set up monitoring dashboards

The memory schema

What KAI remembers between calls.

→ The data model
→ The API contract
→ The infra config
→ The monitoring config

The brief

A real demo brief.

"Stand up a multi-tenant SaaS API on Cloudflare Workers. PostgreSQL via Hyperdrive. R2 for assets. Queues for async work."

Below is KAI's actual execution trace for this brief — every step, every MCP call, every observation, every audit trail entry.

The execution trace

What KAI did.

kai@vzu-os — trace.log

01 ◉ kai > read brief
02 ◐ kai > thinking: multi-tenant saas. cloudflare workers + hyperdrive. 8-12 weeks.
03 ◇ kai > mcp.db.schema > tenants, users, organizations, plans, subscriptions, events
04 ◇ kai > mcp.db.migrate > applied 12 migrations to staging
05 ◇ kai > mcp.api.deploy > hono + trpc + zod, deployed to workers
06 ◇ kai > mcp.queue.worker > queues: email, billing, webhooks, exports
07 ◇ kai > mcp.infra.terraform > r2 buckets, hyperdrive bindings, d1 for flags
08 ◇ kai > mcp.monitor.grafana > dashboards: requests, errors, latency, queue depth
09 ⏵ kai > ship. 10 weeks. 47 services. 99.97% uptime.

kai>

Audit trail · 7 entries

T+00:00:00 brief.received
T+00:01:30 architecture.drafted
T+00:04:00 data.model.finalized
T+02:00:00 db.migrations.applied
T+05:00:00 api.deployed.staging
T+08:00:00 infra.applied
T+10:00:00 brief.completed

The numbers

KAI, by the numbers.

10 weeks

Median brief duration
47

Services deployed
99.97%

Uptime
12

Migrations

The other 11

Every agent in the family.

Get in touch

Brief KAI.

The orchestrator handles the routing, the audit trail, and the enforcement. KAI handles the work. The operator owns the goal.

Brief VZU → stack.vzu.ca →