Brandon-Action Items¶

Updated: 2026-06-06

Purpose¶

This is the consolidated list of items that need Brandon Tomkins's specific input or decision. Most are non-urgent. The list exists so items don't disappear between conversations.

When Brandon resolves an item, it moves to the Decision Log (if a decision was made) or a DDM (if deferred), and gets removed from this list.

Money decisions¶

SWTR / new-entity bootstrap (gates the whole rental business)¶

What's needed: Brandon-only actions required before SWTR can transact. Per doc 10 §A of the since-Friday batch: - [ ] File STR EIN (IRS website) - [ ] Register "Superior Water Truck Rental" d/b/a under STR - [ ] Open STR bank account under new EIN - [ ] Apply for STR Arizona TPT license (Personal Property Rental; + Transporting if billing transport directly) - [ ] Bind insurance (fleet, operator liability, transit) — with broker - [ ] Enter live Stripe credentials / accept Stripe terms (never done by agents)

CPA/attorney confirmations (cheapest to get right early): - [ ] Two-layer TPT treatment confirmed in writing (STE→STR resale-exempt; STR remits retail) - [ ] Self-rental / bonus-depreciation interaction modeled before electing depreciation - [ ] FMV master-lease rate confirmed defensible - [ ] STE entity election confirmed (pass-through, not C-corp) - [ ] AZ attorney reviews master lease + rental agreement together - [ ] Insurance broker confirms policy forms for unregistered off-road rental equip

Brandon-decisions the build is waiting on: - [ ] Starting fleet — how many 2k / 4k rebuild-ready trucks to seed listings - [ ] Domain — confirm superiorwatertruckrental.com OR commit to cactuswatertruckrental.com (CWTR rename pending per notepads/cwtr-rename-pending.md) - [ ] Phone — new line or routed extension - [ ] Delivery — DET as haul arm, or third-party (affects transport TPT + insurance) - [ ] Deposit policy — flat / % of monthly / risk-based - [ ] CRM rollout — SWTR-first then absorb others, or migrate existing table day one

Why it matters: Every one of these gates either the legal ability to operate (EIN/TPT/insurance), the payment rails (Stripe), or the build sequencing (CRM/domain). The substrate can't ship the website until shared CRM + Supabase/auth lands, and that can't land until the CRM-rollout decision lands.

SWTR legal documents — drafted, but Word files not in framework¶

What's drafted (per doc 05 of since-Friday batch): - SEQ Water Truck Rental Agreement (.docx, AZ governing law, branded) — final version: turns on broad principle of full renter accountability for ANY damage (§7), with narrow §7.4 exceptions (wear/tear + documented pre-existing + SEQ gross negligence) that make the broad clause enforceable - SEQ Water Truck Inspection Record (.docx, Exhibit A) — 7 sections, makes §7.4 pre-existing-condition exception work as a release gate

Missing from framework: neither .docx is in D:\claude_www\ — only the recap (doc 05) is. Same pattern as Loop artifacts.

Action when located: drop the .docx files in D:\claude_www\.tmp\ or Downloads; mc1-code stages into projects/swtr/legal/ (new subdir alongside docs/ since .docx doesn't render in MkDocs).

Operational hard rule from doc 05: signed check-out inspection + insurance certificate are RELEASE GATES — hard steps in the release process. An unsigned inspection record sinks the whole damage-protection design. (Worth baking into the operational playbook once the artifacts land.)

Still to draft (the next legal document): - Master lease (STE → STR) — wholesale layer: fleet master lease, FMV rate schedule, maintenance/insurance allocation, explicit sublease permission. Should be reviewed alongside the rental agreement by the same AZ attorney.

Loop (task aggregator) — red-tag blockers + missing framework artifacts¶

What it is: Loop — substrate-native triage-first task aggregator built through v0.2 prototypes on 5/31. The "one list of what needs doing" fed by agents + Gmail with an Inbox → Keep/Snooze/Dismiss gate. Detail in projects/enterprise/strategy/archive/since-friday-2026-05-29/03-loop-task-management.txt.

Two red-tag blockers (Brandon-only, gate the email-ingestion half going live): - [ ] Gmail OAuth (grant the Loop app access to Brandon's Gmail) - [ ] Decide which inboxes Loop watches (full list or curated subset)

Three artifacts that don't appear to be in the framework filesystem: - MC1-handoff-20260531-loop-task-app.md — the 5/31 build handoff (referenced in doc 03 but not found at handoffs/) - loop-v0_2-prototype.html — the v0.2 build target (referenced but not found) - The earlier loop-v0 + loop-v0_1 HTML prototypes

Action item for routing once you locate them: drop them in D:\claude_www\.tmp\ or Downloads + mc1-code stages into projects/loop/ (substrate-native build, would also need a workspace + mailbox stub similar to MC-NPF infrastructure).

Why it matters: Loop is the to-do system on phone + desktop. Without the red tags resolved, the email-ingestion half is offline + Loop only sees agent-surfaced tasks (not Gmail). Without the framework-resident artifacts, the prototype work isn't persisted across machines/sessions.

Substrate agent cost caps¶

Status: Effectively superseded 2026-05-27 by DECISION-LLM-INFRA-MANAGED-API (managed API + measure-then-decide). The $20/agent + $100/aggregate caps remain as runaway guardrails, but the real decision input is now measured monthly spend vs the $200/mo trigger line. Stand up the fleet on tiered API (Opus/Sonnet/Haiku per role), measure one month, then decide self-hosting question. Open piece: Brandon may want to explicitly confirm the $20/$100 guardrails as ceilings (separate from the $200/mo measure-trigger). Why it matters: As more managers migrate to substrate-native (OM next), runaway protection still needs explicit policy even though the strategic cost question is settled.

Load board subscription approval¶

What's needed: Approval for DAT or Truckstop.com subscription (~$150-450/month) Why it matters: Unlocks Phase 1 of DET dispatch capability (load-board read-only)

Email infrastructure cost¶

What's needed: Confirm Google Workspace vs. Microsoft 365 vs. existing arrangement Why it matters: Stage 1 of AI Communication Architecture needs the underlying email infrastructure

Shared services (build specs awaiting Brandon resolution)¶

ATG Inventory Service — five open items in §9 of the spec¶

The build spec is filed at Methodology · Service Specs · ATG Inventory Service. The builder will not guess on these; resolve before code:

🔴 Availability-state source (the gating field) — Excel column added to MASTER vs. small state-edit tool. This is the most important field; it gates every site.
🔴 Cost source — Excel master vs. QB. Confirm where acquisition + rebuild costs live.
🔴 Sync mechanism + cadence — recommend scheduled + on-demand Excel import (keeps Brandon's current workflow unchanged while websites get a live source).
🟡 Co-host with the central CRM — recommend yes (same box/pattern, separate service + DB, separate fate from the strained seq-prox pool).
🟡 Price fields now vs. when pricing engine lands — recommend: build the prices table + public field now, populate it when the pricing logic is built next step.

Why these matter: Each is named in §9 of the spec as a "resolve before build" gate. The service is the inventory analog of the central CRM — three websites (rental, parts, sales) will plug into its single contract instead of building their own lookups. The cost/price boundary (§3) is non-negotiable; the other items resolve where data flows from.

Operational clarifications¶

CTP /admin feature gap list¶

Status: ~~Brandon-action~~ Reassigned to WDM 2026-05-26 per DECISION-CACTUS-IMS-WDM-OWNS. WDM enumerates the gap from logged-in /admin inspection, scopes the port, sequences the work. Brandon out of the loop unless WDM surfaces a four-test failure. SE-App stands by for coordination ping. Why it matters: Was blocking the Cactus IMS → cactustruckparts.com merge; no longer blocked.

Team roster verification¶

What's needed:

Who actually operates the CTP parts photo capture? (Was attributed to Karin, who is cleaning staff — correction 2026-05-27)
Who handles day-to-day QB/accounting work?
Is there a bilingual operational bridge to Jose's crew, or does Brandon handle directly?
1099 crew names verification against payroll (Joel Baca, Albert Moran, Luis Roa, Mario Ciriaco, plus others)
James (DET driver) — employment arrangement and current readiness
Employment relationships across the team (which entity, W-2 vs. 1099/contractor)

Shop/office operational baseline capture — drives substrate requirements (5/31 next-session lead)¶

What's needed: A structured capture of what the shop and office actually do day-to-day — workflows, hand-offs, tools used, decisions made, where time goes. Per doc 07 reframing (since-Friday archive doc 07), this is the right research artifact that early apps like YMS/TruckBid/HD-pricing were learning toward. It is NOT a "migration ledger" — the substrate doesn't migrate the old code, it rebuilds from operational requirements.

Why it matters: The substrate-native manager-agent rollout sequencing (OM after Phase 0, future SP/CTP/SWTR operational tooling, AMP build-out) depends on knowing the actual operational shape — not the imagined one. Without this, manager-agents get built on inferred requirements and miss the real seams.

Scope: - Shop side: parts intake, teardown, photograph, inventory entry, listing prep, customer pickup, work-order flow, Jose-crew daily rhythm - Office side: phones, email, quoting, order entry, AR/AP touch-points, Andrea's daily flow, Chamine's QB rhythm, customer/vendor coordination

Tied to: Spanish-first directive (DECISION-SHOP-CONTENT-SPANISH-FIRST 5/31) means shop-side capture has UX implications for any tool that comes out of this; office-side has fewer language constraints.

Format: mc1-code-led interview/observation pass, written to projects/enterprise/strategy/research/operational-baseline-{shop,office}.md (NEW paths — not yet created). Brandon answers questions; mc1-code structures and writes. Brandon ratifies the captured baseline; manager-agents inherit it as a requirements doc.

Status: open, NOT scheduled. Worth a dedicated session.

DET regulatory specifics¶

What's needed: Source paperwork from 2026-05-27 filing — exact authority type (interstate vs. intrastate Arizona), insurance coverage limits (auto liability $, cargo $), ELD operational status Why it matters: Documentation accuracy and broker-requirement verification for load-board work Timing: Brandon mentioned providing tomorrow (2026-05-28)

Ledwell trailer equipment type¶

What's needed: Confirm flatbed / step-deck / RGN Why it matters: Determines what loads DET can carry on the load boards

China supplier relationship for SP¶

What's needed: Single supplier vs. multiple, shipment scale, direct-import vs. through-distributor, quality consistency Why it matters: SP's competitive moat (and SP-MX expansion case) depends on supplier reliability

Notepad/routing-prompt deployment status¶

What's needed: Which routing prompts from chat-MC1 conversations have been pasted into MC1-code? Why it matters: chat-MC1 has drafted many directives; the framework only reflects them if MC1-code received and executed them. Catching up routing-prompt deployment is the bridge between conversation and framework reality.

Strategic conversations needed¶

CPA structural review¶

What's needed: Schedule a multi-entity-specialized CPA for a strategic structural review Topics: S-Corp election / QSub structure analysis for the ATE-affiliated INC group (Form 1120 consolidation question), arm's-length transfer-pricing documentation for inter-entity transactions (especially DET → STE, a related-party transaction), NewPath 501©(3) firewall verification, Tamarack capitalization mechanics, Section 199A capture, Arizona TPT optimization + combined reporting, estate planning, SJB → TCR capital-flow mechanism (intercompany loan vs ATE distribution / recapitalization) Bring: Entity Structure & Reasoning, the Tax Analysis — open angles for CPA review, and Parked Questions for MC-ATG (~60 questions across all entities)

Dennis interviews (Salt Lake City microphone) — NPF SoT scope, ATG-relevant context only¶

What's needed: Schedule first Dennis interview Why it matters: Dennis is a consultant to ATG; his strategic context lives in the NPF SoT (his own enterprise: partner-side entities, real-estate, reentry programming, cross-enterprise capital framework). Interview capture is to refresh ATG's understanding of where Dennis's consulting lens sharpens ATG strategic thinking — not to document partner-side detail on ATG site. Resources: Microphone installed at Salt Lake City NewPath office

Jose interview planning (Phoenix microphone)¶

What's needed: Phoenix microphone setup; Spanish-language interview structure; conversation with Jose about the framing Why it matters: Jose's 16 years of shop-floor judgment + his life context need capture in his primary language

Dennis-Brandon conversation on framework-as-product¶

What's needed: Joint ownership structure, on-site-server rationale, model-aggregation consent model, target customer profile, who does the work (capacity question) Why it matters: Framework-as-product strategic intent needs principal alignment before any productization work begins

Dennis-Brandon consulting-relationship documentation¶

What's needed: Documented consulting/joint-arm agreements where ATG and Dennis's enterprise intersect (framework-as-product, any joint deals). Cross-enterprise capital framing (Tamarack, the cross-enterprise reentry thesis) lives in the NPF SoT, not in ATG action items. Why it matters: Long-standing consulting relationship + any joint-arm intersections need written succession / dispute structure on the ATG side; NPF SoT carries the partner-enterprise side.

Risk and compliance¶

What's needed: Tax counsel review of NewPath relationships with commonly-controlled for-profit entities. MC-NPF specifically flagged exposure 2026-05-29 around two IRS doctrines: private benefit (benefits flowing improperly to private interests) + inurement (insider economic benefit). Both can threaten 501c3 status.

Three priority doc categories MC-NPF needs to review on code-access grant: 1. Governance + formation docs (Form 1023 application, IRS determination letter, bylaws, conflict-of-interest policy, board minutes) 2. Related-party transaction records (NPF↔Tamarack lending arrangements, NPF↔Thorson construction labor contracts, NPF↔ATG-family employment arrangements) 3. Program service history (housing units placed, reentry participants served, outcomes data)

Two parts to this: - (a) Provision MC-NPF code-access (filesystem access to NPF docs) so MC-NPF can do the document-review work - (b) Schedule the 501c3-specialist tax counsel review — this is the human-attorney layer; MC-NPF does the document analysis, attorney does the legal opinion

Why it matters: 501©(3) status is critical infrastructure (the whole reentry-housing thesis depends on it); exposure is unforced-error-shaped (papering arm's-length relationships is cheap to do early, expensive to retrofit if audited)

MC-NPF code-access provisioning¶

What's needed: Provision MC-NPF (NewPath Foundation strategic consultant) with mc1-code-style filesystem access so she can read NPF governance + transaction + program-service docs, and write findings to her workspace + mailbox.

Status as of 2026-06-01: mc1-code created stub workspace at projects/mc-npf/ + mailbox stubs at agent_comms/to_mc-npf/ + from_mc-npf/. Per doc 09: "MC-NPF code access was forthcoming" — pending Brandon's provisioning step.

Why it matters: Without code-access, MC-NPF can only do the 501c3 review in chat — slower, doesn't persist findings durably, no audit trail.

Email Stage 3 autonomous response risk tolerance¶

What's needed: What pattern of customer-facing autonomous AI response feels acceptable? Why it matters: Stage 3 of email architecture depends on Brandon's risk tolerance threshold

STR rental-fleet truck ownership¶

What's needed: Which entity owns the trucks STR rents — SEQ (direct from build), STE (held finished inventory leased to STR per the SWTR master-lease draft), or a hybrid by category / point-in-lifecycle?

Why it matters: Tax + balance-sheet treatment differs materially across the two separate INCs. The SWTR master-lease draft currently names STE as the truck-OWNER leasing to STR, but the canonical entity definition (STE holds finished inventory and sells; SEQ builds) leaves ownership-at-rental ambiguous. Flagged 2026-06-06 by MC-ATG via OM on entities/str.md.

Living document. Items resolve into Decision Log or DDMs. Last updated 2026-06-06 by claude-operations-manager (STR ownership flag, per MC-ATG directive).