Umbra Group / Studio / Studio Ops
v1.0 · Internal
DocumentStudio Ops — Master
ClassificationInternal
Versionv1.0 · 2026.04
OwnerUmbra Studio
Studio Ops Kit · Master Operating Framework · Internal

Studio Operating Framework.

The machine that runs the Studio. This is the master document for how Umbra Studio itself operates — the tools we use to build agent systems, the infrastructure that runs them, the hardware we run it on, the security posture around client data, and the internal discipline that makes the Lighthouse Sprint reproducible.

Paired with five operational sub-documents — Agent Runtime, Collaboration Stack, Studio Internals, Hardware & Workspace, Security & Secrets. Together they describe everything it takes to deliver a Lighthouse Sprint without reinventing the wheel. Read before the first sprint; reviewed quarterly.

§01

How to use this kit.

Audience · reading path · relationship to the Lighthouse Kit

The Studio Ops Kit is the mirror image of the Lighthouse Kit. The Lighthouse Kit is what we sell — the playbooks a Sprint Lead executes inside a client engagement. The Studio Ops Kit is what makes selling that possible — the tools, infrastructure, hardware, and internal discipline that let the Studio deliver a Lighthouse Sprint reliably, profitably, and without the operating model being a bottleneck.

This master document is the map. Read it first. It establishes the operating principles, the full stack map, the budget envelope, the governance we apply to ourselves, and which of the five sub-documents to reach for when a specific question comes up.

If a single person is running the Studio, this kit is the operating manual for that person. If the Studio scales past one, this kit is what onboards the second, third, and fourth person without everything that one person knows having to be re-learned by tribal osmosis.

Who reads what

RoleMust readShould readReference when needed
Founder / PrincipalAll sixAll six
Sprint LeadMaster, Collaboration, InternalsAgent Runtime, SecurityHardware
Agent EngineerMaster, Agent Runtime, SecurityHardwareCollaboration, Internals
Governance ArchitectMaster, Agent Runtime, SecurityInternalsHardware, Collaboration
New hire (day 1)MasterTheir role’s “Must read”Everything else

Relationship to the Lighthouse Kit

Every client-facing choice in the Lighthouse Kit is underwritten by a studio-side capability documented here. When the Discovery Playbook says “the Agent Engineer instruments monitoring”, it is the Agent Runtime doc that defines what “instrumented” actually means in our stack. When the Handoff Playbook says “transfer credentials via a secure vault”, it is the Security doc that specifies the vault, the rotation cadence, and the revocation procedure.

If you ever find a gap — the Lighthouse Kit asks a Sprint Lead to do something that the Studio Ops Kit does not enable — that is a P0 operational bug. Fix it in the Ops Kit before the next sprint starts.

§ Principle
What we sell runs on what we have. The Studio Ops Kit exists because we refuse to ship promises the Studio cannot keep. Every capability in the Lighthouse Kit should have an owner, a tool, a budget, and a runbook here.
§02

Operating principles.

Seven commitments that shape every tooling, infra, and hardware decision

Tooling decisions made in isolation drift into a zoo. These seven principles are the constitution: every sub-document in this kit should defer to them. When a specific recommendation in a sub-document conflicts with a principle here, the principle wins and the sub-doc is wrong.

P-01

Own the primitives, rent the commodities.

We own prompt logic, pattern library, agent specs, governance runbooks, and the state-machine shape of every workflow. We rent everything else — compute, storage, messaging, identity, billing. If a vendor raises prices 3× overnight, the switching cost on rented commodities should be measured in days, not quarters.

P-02

Spreadsheet-first, database when pressure demands.

USP-009 applies to the Studio, not just to client systems. Every new workflow starts as a Google Sheet with Apps Script glue. Graduate to a real database (Supabase / Airtable / Postgres) only when a named pressure forces the move: concurrent writes, row count past ~5k, sub-second query needs, or a compliance requirement. Never migrate on aesthetics.

P-03

One model, three tiers, measured per run.

Default to Claude. Use Opus 4.6 for planning (ICPs, spec writing, post-mortems), Sonnet 4.6 for execution (most production agent runs), Haiku 4.5 for high-volume tight-loop tasks (classifiers, routers, format-check guards). Every agent run logs token count and dollar cost to the state sheet — no exceptions.

P-04

Governance is not a tool — it is a column.

No monitoring tool survives contact with a sprint if it lives in a separate UI the client will never check. Every governance signal must land on the state sheet (severity, last-alert, last-rollback, last-audit-review, readiness). The state sheet is the dashboard. External observability platforms are nice-to-have, not core.

P-05

Reversible by default.

Every tool choice, every deployment, every credential must have a documented “how to back out” runbook. If you cannot describe the reversal in under 15 minutes of work, the choice is too expensive. This principle is why we avoid bespoke infrastructure for anything that has not yet proven itself across three sprints.

P-06

Client data never lives on a laptop.

Client credentials are stored in the vault. Client data is read from the client’s systems in place. No client data is persisted to a personal machine. Local caches are ephemeral and wiped at end-of-sprint. This principle dictates many hardware and workspace choices.

P-07

Every Friday, the Studio ships a demo to itself.

USP-011 applies internally. We run a Friday demo of any Studio-internal tooling change, pattern addition, or process revision — in a 30-minute calendar block, solo or with collaborators. No internal improvement lands on a Monday without a Friday demo. This keeps the Studio’s own operating system from rotting.

§ Counterweight
These principles exist to resist drift, not to resist thought. When one genuinely conflicts with the sprint in front of us — escalate, document the exception, and update the principle in the next quarterly review. Never silently violate.
§03

The stack map.

Five layers that describe every tool, service, and piece of hardware the Studio depends on

We describe the Studio stack in five layers, top to bottom. Every tool mentioned in any sub-document of this kit maps to exactly one layer. If it does not map cleanly, it is probably redundant with something that already does.

L5 · Surface
Client-facing surface.
Web properties (umbragroup.dev, studio.umbragroup.dev, lighthouse kit, case studies), outreach emails, sales deck, invoices, contracts. Everything the client sees.
Vercel · Gmail · Stripe
L4 · Collaboration
Collaboration & comms.
Where the Studio and the client talk, track, and hand off. Project tracking, async demos, design files, calendars, shared docs. Sub-doc: Collaboration Stack.
Linear · Slack · Figma · Loom
L3 · Build & Runtime
Agent build & runtime.
Where agents are authored, versioned, deployed, and run. Claude API, Agent SDK, MCP servers, state sheets, deployment targets, observability. Sub-doc: Agent Runtime.
Claude · GitHub · Vercel · Cloud Run
L2 · Studio Core
Studio internals.
The Studio’s own business: CRM, pipeline, invoicing, contracts, knowledge base, pattern library, post-mortem repository. Sub-doc: Studio Internals.
Sheets · Notion · Stripe
L1 · Physical
Hardware & foundation.
Workstations, monitors, backup, network, identity (1Password, Google), secrets vault, physical workspace, travel kit. Sub-docs: Hardware, Security.
MacBook · 1Password · Google

What lives where, at a glance

CapabilityLayerPrimary toolAlt under trade-off
Agent reasoningL3Claude API (Opus/Sonnet/Haiku)Local model (Ollama) for cost-sensitive routing
Agent orchestrationL3Claude Agent SDKRaw API + custom harness
External data / toolsL3MCP serversDirect API clients in code
Workflow stateL3Google Sheets + Apps ScriptAirtable · Supabase (when USP-009 pressure fires)
Code hostingL3GitHub (private repos)GitLab (if EU residency requirement)
Agent deploymentL3Vercel Functions · Google Cloud RunFly.io · Render
Web hostingL5VercelCloudflare Pages
Project trackingL4LinearNotion · GitHub Projects
Client commsL4Slack Connect · GmailMicrosoft Teams (client-led)
DesignL4FigmaPenpot (open-source alt)
Async demosL4LoomTella · Descript
Knowledge base / docsL2Notion workspaceGoogle Docs folder tree
CRM / pipelineL2Google Sheet (CRM tab)Attio · Folk (when past ~40 deals)
InvoicingL2Stripe InvoicingWave · QuickBooks
Contracts / NDAsL2DocuSignHelloSign · PDF + email
Secrets vaultL11Password TeamsBitwarden · Doppler (for infra secrets)
Identity / SSOL1Google Workspace
WorkstationL1MacBook Pro M4 Max (primary); Mac mini M4 (desk)Framework laptop (Linux, long-term optional)
BackupL1Time Machine to external SSD + BackblazeArq to B2
§04

Budget & cost model.

Monthly envelope · per-sprint variable · capex cadence

The Studio runs on a two-line cost model: a monthly fixed envelope that holds whether or not a sprint is live, and a per-sprint variable line that tracks with engagement intensity. Numbers here are working targets at v1.0 — they will move, and the shape of the model matters more than the specific dollar figures.

Monthly fixed envelope — what it costs to keep the Studio running with zero sprints active

CategoryTool / line itemMonthly (USD)Notes
L3 Runtime baselineClaude API (standby/dev)~$80Dev iteration, internal agents, small invocations.
Vercel Pro$20All web properties — studio, lighthouse, case studies.
L4 CollaborationLinear (Standard)$10Single-seat baseline.
Slack (Pro)$9For Slack Connect with clients.
Figma (Professional)$18Files persist; minimal editor seats.
L2 Studio CoreNotion (Plus)$12Knowledge base, pattern library, post-mortems.
Google Workspace (Business Starter)$7Identity, mail, Sheets = state-machine substrate.
DocuSign (Personal)$15Contracts, SOWs, NDAs.
L1 Foundation1Password (Teams Starter)$20Secrets vault for all categories.
Backblaze (unlimited)$9Continuous backup.
Domain registrations~$4Amortized annually.
ReserveBuffer / replacements$40Software trials, ad-hoc services.
Envelope total~$244Zero-sprint steady state.

Per-sprint variable line — what an active Lighthouse Sprint adds

PhaseVariable cost driverExpected (USD)Notes
Discovery (W1–2)Claude API (planning-heavy, Opus-biased)$120–$200ICP writing, spec drafting, stakeholder synthesis.
Redesign (W3–4)Claude API + light MCP calls$180–$300Iteration on agent specs and wrapper design.
Build (W5–9)Claude API (Sonnet heavy) + compute$900–$1,600Production agent runs, ramp to 300+ invocations/wk.
Handoff (W10–11)Claude API + training rehearsals$200–$350Full-day rehearsal, supervised operation.
Support (30d)Claude API (spot checks)$60–$120Light ad-hoc invocations during support window.
Expected total$1,460–$2,570COGS for a single sprint at v1.0 scale.
§ Margin check
A Lighthouse Sprint priced at the low end of the qualified range should leave at least 15–20× the variable COGS as contribution margin. If the variable line creeps above ~10% of sprint price, stop and diagnose — that usually means an agent is running without a cost cap.

Capex cadence — hardware and one-time investments

ItemExpected cadenceTarget spendDriver
Primary workstation refreshEvery 3–4 yrs$3,500–$4,500Dev speed & multi-core agent test harnesses.
Secondary workstation (desk Mac mini)Every 4 yrs~$1,500Always-on backup runtime & build host.
Monitor (4K, 27”+)Every 5 yrs~$700Screen real estate for state sheets + code + demo.
External SSD (2TB+)Every 3 yrs~$200Time Machine + sprint scratch.
Mobile test devicesOpportunisticvariesPhones for mobile web QA; keep one low-spec Android.
Network (router, mesh)Every 5 yrs~$400Stable uplink during demo calls.
§05

Studio-level governance.

The same six components — turned inward

We apply the Six Governance Components from the Lighthouse framework (USO-LH-07) to the Studio itself. Same vocabulary, different scope. This section maps each component to its Studio-internal expression so that if a Studio-internal tool, pipeline, or process fails, we already know where to look.

ComponentClient scope (Lighthouse)Studio scope (this kit)Where it lives
MonitoringAgent throughput, error rate, latency per runSprint cadence, pipeline health, renewal risk, tool cost driftStudio State Sheet · Linear cycle view
AlertingSEV-1/2/3 incident alerts on client agentsCost over budget, credential expiring, backup missed, renewal dueCalendar reminders · email rules · budget sheet triggers
Audit trailPer-invocation log of every agent runDecision log of every tool swap, pattern promotion, sprint variationstudio-decision-log sheet
Escalation3-tier ladder to client sponsorCollaborator or peer review for anything touching contracts, pricing, IPPeer review checklist · founder final sign-off
OverridePause, resume, force-failure, skip-next, redirectFreeze tooling during an active sprint (no mid-sprint migrations)P-05 reversibility principle · deploy freeze window
RollbackSnapshot, restore, replay, partialTool choice reversal runbook; every subscription has a 15-min cancellation pathSecurity doc §06 revocation · Internals doc §08 tool reversal

Four studio-level SEV definitions

SeverityConditionACK windowResponse
Studio-SEV-1Client data exposure risk, credential leak, active production incident on a client system traced to Studio tooling< 15 minPause affected agents, invoke Security §09 incident response, write post-mortem within 48 h.
Studio-SEV-2Sprint at risk (gate blocker, stakeholder frustration, missed weekly demo), or Studio tooling outage blocking work > 4 h< 2 hTriage and escalate to Sprint Lead; document in Linear. Workaround must exist by EOD.
Studio-SEV-3Tooling degradation, cost drift, backup miss, vendor notice, renewal-due warning< 24 hLog to decision log; resolve in next work session or queue for next quarterly review.
Studio-SEV-0Observation, informational, “nice to look at next quarter”Drop into quarterly review backlog.
§ Inversion
If we would not accept weak monitoring or escalation inside a client engagement, we will not accept it inside the Studio. The Studio is the Studio’s first client.
§06

How the six docs interlock.

What lives where · cross-references · reading order

The Ops Kit is six documents. This master plus five sub-documents. Each sub-doc has a narrow, non-overlapping scope; cross-references are explicit. If you cannot find a topic where you expected it, check the owner matrix in §08 first — it indexes every topic to its home doc.

Cross-reference index — where to find common topics

Looking forPrimary docSection
Which model tier to use for a new agentAgent Runtime§03 Model tiers
How to publish an MCP server for a client integrationAgent Runtime§04 MCP patterns
When to graduate from Sheets to a databaseAgent Runtime§05 State graduation
Slack Connect setup with a new clientCollaboration§03 Client channels
Linear project template for a sprintCollaboration§04 Tracking
Loom recording & retention policyCollaboration§05 Async demos
Where the pattern library livesInternals§05 Pattern library
Invoice schedule & milestone paymentsInternals§03 Billing
Post-mortem repository structureInternals§07 Post-mortems
Workstation spec for a new hireHardware§03 Workstation profiles
Backup & recovery procedureHardware§05 Backup
Credential rotation cadenceSecurity§04 Rotation
Client credential revocation at Gate 4Security§05 Revocation
Data retention & end-of-engagement wipeSecurity§07 Retention
IP ownership on patterns extracted from client workSecurity§08 Contracts & IP
§07

Graduation path & decision log.

When to move a tool up a tier · how we record the move

Tools are added, replaced, and promoted over time. P-05 (reversible by default) means we move with humility — but we still move. This section defines the named pressures that legitimately force a tool graduation, and the discipline for recording the decision.

Five legitimate pressures to graduate a tool

  1. Concurrency pressure. The current tool cannot tolerate the number of simultaneous writers, readers, or runs the workflow now requires. Example: a state sheet with more than ~3 Apps Scripts writing to it at once will race.
  2. Volume pressure. Row count, invocation count, or file size passes a threshold where the current tool becomes slow or unusable. Example: a Sheet past ~5,000 rows in the active tab starts to feel sluggish for interactive use.
  3. Latency pressure. An operation that must complete in under a specific budget can no longer hit it. Example: a governance readiness check that needs to return in under 300 ms.
  4. Compliance pressure. A client or regulatory requirement makes the current tool legally or contractually unusable. Example: PHI, EU residency, SOC 2 auditor ask.
  5. Vendor pressure. The current tool changes pricing, terms, or availability in a way that makes it untenable. Example: an MCP server’s paid tier spikes 5× or shutters.
§ Forbidden reasons
Aesthetics, novelty, because a thread on social said so, because a peer Studio uses something shinier. None of these count as a pressure. If you cannot name which of the five above applies, stop — the move is premature.

Decision log — seven-column format

ColumnPurpose
dateYYYY-MM-DD when the decision was logged.
fromTool / service / setup being replaced or evolved.
toTool / service / setup being adopted.
pressureWhich of the five legitimate pressures applies. Cite one.
cost_deltaMonthly (or one-time) cost difference in USD, signed.
rollbackOne-sentence description of how to back out, and target duration.
ownerWho made and owns the decision.

The decision log lives in the studio-ops Google Sheet, tab decision-log. Every Friday demo that changes tooling appends a row. The quarterly review (§09) reads the last quarter’s rows first — that is where pattern drift becomes visible.

§08

Owner matrix.

Who owns every layer · who steps in when an owner is unavailable

Every tool and process in this kit has exactly one primary owner. When the Studio is one person, the primary owner is the founder across the board — but we still record it, because the matrix is what makes the second hire onboard in a week instead of a month.

AreaPrimary ownerBackupReview cadence
Operating principles & stack mapFounderSprint LeadQuarterly
Budget envelopeFounderMonthly
Claude API usage & cost capsAgent EngineerFounderWeekly (during active sprint)
Agent SDK & prompt versioningAgent EngineerFounderPer-sprint
MCP server inventoryAgent EngineerFounderQuarterly
State-sheet architectureAgent EngineerGovernance ArchitectPer-sprint
Deployment targets (Vercel / Cloud Run)Agent EngineerFounderQuarterly
Observability & dashboardsGovernance ArchitectAgent EngineerPer-sprint
Client Slack Connect channelsSprint LeadFounderPer-sprint
Linear workspace & templatesSprint LeadFounderQuarterly
Figma librarySprint LeadFounderQuarterly
Sales pipeline / CRMFounderSprint LeadWeekly
Invoicing & billingFounderMonthly
Contracts, SOWs, NDAsFounderPer-sprint
Pattern library (USP-001…014)FounderAgent EngineerPer-sprint + quarterly
Post-mortem repositoryGovernance ArchitectFounderAfter each incident
Workstations & hardwareFounderSemi-annual
Backup & restore testFounderMonthly restore-test
Secrets vaultFounderMonthly rotation audit
Client credential lifecycleFounderSprint LeadPer-sprint + at Gate 4
Incident response (studio-side)FounderGovernance ArchitectDrill semi-annually
§09

Quarterly review cadence.

Four-hour working session · standing agenda · what ships out the other side

Once a quarter, the Studio stops client work for a full half-day and reviews its own operating system. This is non-negotiable — it is how we resist the drift that eventually poisons any Studio’s tooling stack. Same day each quarter; same agenda; same deliverables.

Standing four-hour agenda

BlockDurationWhat happens
B1 · Prior-quarter read-out30 minRead every row of the decision log since last review. Circle anything that violated P-01–P-07 without an explicit exception.
B2 · Cost audit45 minReconcile actual vs. envelope & per-sprint variable. Flag any line that moved > 15%. Kill one tool if possible.
B3 · Stack map sweep45 minWalk the §03 stack map row by row. For each row, does the primary tool still fit? Has the named alt become more attractive? Is anything orphaned?
B4 · Pattern library review30 minReview USP-001…N. Any new candidates from the last quarter? Any pattern that failed in reuse? Retire or demote as needed.
B5 · Post-mortem sweep30 minRead every studio-side post-mortem written this quarter. Extract one remediation into the next quarter’s work queue.
B6 · Backup & restore drill30 minLive restore of one arbitrary file from last month’s backup. If restore fails or takes > 10 min, Studio-SEV-2 until resolved.
B7 · Next-quarter plan30 minPick exactly 2–3 things to change next quarter. Never more.

Deliverables out the other side

  1. Updated MEMORY index in Notion workspace with the quarter’s changes and rationale.
  2. New decision-log rows for every change adopted in B3/B4/B7.
  3. Budget sheet reconciled — variance rows resolved or moved to the next quarter.
  4. Two or three targeted changes scheduled for the next quarter — no more, each with an owner and due date.
  5. One retired tool — ideally. Subtract before adding.
§ The subtraction rule
Every quarter should end with at least one thing removed from the stack. Without this discipline, complexity accumulates silently. A quarter with only additions is a warning sign — ask what is secretly breaking.
§10

Studio Ops checklist.

What the Studio operator verifies, when

Weekly (every Friday, 20 min)

  1. Budget-vs-actual scan — any category drifting past 15% of envelope?
  2. Sprint state sheet review — governance columns green for every active sprint?
  3. Backup verification — last Time Machine & Backblaze timestamps healthy?
  4. Inbox zero on client comms — no thread older than 48 h without a response.
  5. Friday demo — any internal tooling change since Monday? Record one even if solo.

Monthly (first working day of the month, 60 min)

  1. Reconcile Claude API invoice against expected per-sprint variable lines.
  2. 1Password monthly rotation audit — any item older than its rotation target?
  3. Restore-test: pick a random file backed up in the last month — restore it. Must work in < 10 min.
  4. Pattern library sweep — any new candidate patterns from the month? Any reuse failure?
  5. Pipeline review — any deal stuck in the same stage for more than 30 days? Either advance or close-lost.
  6. Invoicing run — generate, send, and record expected receipt dates.

Quarterly (§09 cadence)

  1. Execute the seven-block agenda in §09 end to end.
  2. Retire at least one tool.
  3. Post the updated stack map to Notion.
  4. Schedule the next quarterly review on the calendar before closing the session.

Per-sprint (at the boundaries)

  1. Sprint start: verify every tool for each layer is reachable; credentials healthy; budget envelope adjusted for this sprint’s variable line.
  2. Gate 4: client credentials revoked per Security §05; post-mortem filed if any SEV-2/3 occurred; patterns nominated for the library.
  3. Support window close (30 d post Gate 4): retention policy executed per Security §07; final variable-cost reconciliation.
§ Bottom line
A Studio that keeps this checklist will not have a crisis — at worst it will have an inconvenience. A Studio that skips it will, eventually, have all of its fires at once. Pay the small, predictable tax every week.