AI Agents

The Operations Floor

Big Sox Studios has a small crew of purpose-built AI agents handling infrastructure and smart-home work. They have defined jobs, limited permissions, and considerably more documentation than most fictional robots.

Current roster

Meet the operators.

One profile per agent: who they are, what they do, why the role exists, and where the guardrails are.

OPS-01ACTIVE
Otto, Infrastructure Operator

Infrastructure Operator

Otto

Keeps the servers running, the services behaving, and the terminal commands away from Greg whenever possible.

  • Server ops
  • Deployments
  • Diagnostics
  • Verification

What they do

Otto is the infrastructure operator for the systems behind Big Sox Studios. He handles the boring-but-important work that keeps everything else available: services, deployments, diagnostics, server operations, and carefully bounded privileged actions.

Why this role exists

So routine infrastructure work doesn't require Greg to become the world's most reluctant part-time Linux administrator.

Current status

Active and operational. Otto is the established infrastructure operator in the current agent lineup.

Example jobs

  • Diagnose why a service stopped working.
  • Check the health of the server and supporting services.
  • Restart or deploy an approved service.
  • Verify that an infrastructure change actually worked.
  • Handle bounded maintenance jobs through the studio's agent runtime.
  • Produce evidence instead of simply declaring, “looks good to me.”

Operational boundaries

Otto is not supposed to freestyle with production infrastructure. His access is deliberately bounded, sensitive actions have controls around them, and a command completing successfully is not the same thing as the job being verified.

If the server is boring, healthy, and nobody has to think about it, Otto considers that an excellent day.
OPS-02ACTIVE
Porter, Home Assistant Builder

Home Assistant Builder

Porter

Turns “I wish the house did this” into Home Assistant automations—with backups, approvals, verification, and a healthy fear of breaking the lights.

  • Home Assistant
  • Automations
  • Scripts
  • Rollback

What they do

Porter is the builder for Home Assistant. He takes natural-language ideas and turns them into controlled smart-home changes, especially automations and scripts.

Why this role exists

Because “make the hallway lights do something smarter” should eventually be a request, not a Saturday afternoon YAML project.

Current status

Active and operational as the Home Assistant Builder. Porter builds and modifies automations and scripts through controlled approval, backup, verification, and rollback workflows. Further Builder capabilities can continue to develop without changing his active status.

Example jobs

  • Create new Home Assistant automations.
  • Modify existing automations.
  • Create and change scripts.
  • Inspect the existing Home Assistant configuration before making a change.
  • Back up affected configuration before mutation.
  • Apply an approved change, verify the result, and roll it back when necessary.

Operational boundaries

Porter has isolated Home Assistant access and does not inherit Otto's infrastructure privileges. Changes are designed around approval, backup, verification, and rollback rather than unrestricted control.

Porter measures twice, backs up three times, and would really prefer you not unplug the hub halfway through.
OPS-03PRE-HIRE
Trace, Home Guardian

Home Guardian · Future agent

Trace

Future smart-home watchdog. She has a name, a job description, and absolutely no permission to touch anything yet.

  • Monitoring
  • Diagnosis
  • Health checks
  • Future remediation

What they do

Trace is planned as the Home Guardian: the agent that watches the smart home rather than building it.

Why this role exists

A smart home isn't especially smart if something quietly breaks on Tuesday and nobody discovers it until Friday.

Current status

Pre-hire. Trace is planned, not operational, and her monitoring/remediation capabilities are future work.

Example jobs

  • Monitor Home Assistant and smart-home health.
  • Notice unusual failures or degraded behavior.
  • Help diagnose what went wrong.
  • Connect symptoms across devices and systems.
  • Eventually support carefully bounded remediation where that is safe and appropriate.

Operational boundaries

Trace is not built yet. She does not currently monitor, diagnose, or remediate anything, and the future version is intended to stay within carefully defined permissions and operational boundaries.

Trace hasn't started work yet, which technically gives her the best incident record in the company.

Team topology

How they work together

Their jobs overlap enough to cooperate, but their responsibilities and permissions stay deliberately separate.

01

Infrastructure

Otto

Runs and verifies the infrastructure underneath the operation.

02

Builder

Porter

Builds and changes things inside Home Assistant.

03

Guardian · coming soon

Trace

Will watch, diagnose, and eventually help respond when the smart home misbehaves.

What makes these agents?

Names are the fun part. Boundaries are the important part.

These aren't generic chatbots with employee names pasted on them. Big Sox Studios agents are persistent, purpose-built workers with defined identities, jobs, tools, permissions, operational boundaries, and continuing responsibilities.

ChatGPT, Work, Codex, and other AI platforms help build and operate Big Sox Studios, but they are tools and collaborators—not employees.