The investment operating layer where operator-forged rules meet model-agnostic agents. VEX proposes. You approve. Models change; your logic doesn't.
From the Journal
A solo operator loses the one thing that keeps judgment sharp: someone across the desk to say you're wrong. Here's how I rebuild that adversarial loop with AI — first a sparring partner you argue with, then a weekly coach that remembers what you committed to and holds you to it.
Read the journal →Five perspectives on running an investment operation. Conversation proposes; you approve; intelligence surfaces what compounds.
Sense
Market context before you commit — neighborhood posture, comp alignment, and buy-box fit woven into every deal read.
Example skill: forvex-underwriting
Example conversation — draft script
Proposed
Install a Claude Skill and run a real workflow in minutes. Your rules, your approval — not generic prompts.
“weekly accountability”
Weekly accountability check-in for anyone running their own operation — solo operator, founder, freelancer, investor. Opens by holding the user to last week's commitments, works through wins, slippage and blockers, lands 3–5 specific commitments for the week ahead, then emits a carry-forward block so next week's session has memory. Use whenever the user says "weekly accountability", "accountability check-in", "hold me accountable", "weekly review", "weekly debrief", "let's do my week", or starts a new work week. Standalone — no account, no connectors, no database; continuity runs on a plain-text ledger the user keeps.
Download skill“what finishes should I use”
Flip finish & staging selection — picks the finish package (flooring, paint, kitchen, baths, fixtures, curb appeal) that matches a property's comp tier so you maximize ARV without over-improving. Use this skill whenever the user asks "what finishes should I use", "what should I put in this flip", "am I over-improving", "what flooring/counters/cabinets for this house", "finish package", "what level should I renovate to", or gives an ARV/comp tier and wants a materials spec or rehab finish budget breakdown. Standalone and beginner-friendly — runs on user inputs only, no connectors or account required. NOT for personal-home interior design; this is resale-driven.
Download skill“what should I offer”
Maximum Allowable Offer (MAO) calculator for house flips — the most you can pay for a property and still hit your profit target. Use this skill whenever the user asks "what should I offer", "what's my MAO", "max offer", "how much can I pay for this house", "run the 65% rule" or "run the rule of thumb", or gives an ARV and rehab number and wants an offer figure. Standalone and beginner-friendly — runs on user inputs only, no connectors or account required. Also handles "what if" follow-ups on ARV, rehab, margin, financing, and hold time.
Download skillOne operating layer across market context, decision support, execution, capture, and portfolio truth — connected through shared identity, not separate silos.
Every lifecycle touchpoint sharpens the platform — offer, rehab, operate, express. Real operators running real deals forged the rules; the system gets sharper every time you use it.
Not generic prompts — your buy boxes, profiles, and business rules exposed as tools agents can call. Model-agnostic MCP means capability compounds when models improve, not when you rebuild your stack.
Underwriting feeds rehab scope; operational memory grounds content; portfolio truth rolls up to Track — without re-entering context at every handoff.
VEX orchestrates and interprets — it does not invent numbers or publish without you. Deterministic engines run server-side; agents propose; you preview, edit, and approve before anything persists.
Your rules, buy boxes, and approval gates persist regardless of which model or surface you use. Capability changes; your logic doesn't.
When used together, Sense through Track share context and eliminate redundant work. A property analyzed in conversation informs what you see in intelligence.
Your data remains yours. Export capabilities ensure you can always move elsewhere. Integration advantages come from shared context, not walled gardens.
This is intentional design, not accidental architecture. Start with one skill; the operating layer compounds when you're ready to standardize.
forVEX runs on a shared data layer with server-side business rules. Deterministic calculations and workspace data are exposed to AI agents through a model-agnostic tool interface — agents orchestrate, but your rules and approval gates persist regardless of which model you use.
Early access means priority consideration for beta testing and direct input on feature development. We'll reach out as forVEX opens up for your team.
We respect your privacy. Unsubscribe at any time.