> ## Documentation Index
> Fetch the complete documentation index at: https://talunavo.com/llms.txt
> Use this file to discover all available pages before exploring further.
# FAQ
> Is this financial advice? What does it model? Where does my data live? How is this different from a Monte Carlo calculator?
export const FaqList = () => {
const items = [{
q: "Is Talunavo financial advice?",
a: ["No. It models arithmetic under assumptions you can see and change. It never says “you should”; it says “Plan B pays this much less lifetime tax than Plan A under these assumptions.” The decision stays yours, and the export is a document you take to a professional."]
}, {
q: "How is this different from a Monte Carlo retirement calculator?",
a: ["Most calculators return a single success probability and stop. Talunavo compares named plans under identical assumptions, with a full tax model in every year, and treats uncertainty as one of three explicit modes. Monte Carlo, when you run it, returns a distribution with the shortfall’s magnitude and timing, never a single percentage. The store refuses to write one."]
}, {
q: "What does the tax model cover?",
a: ["Nine passes per year: filing status including the transition a death causes, Social Security taxability, ordinary brackets, long-term gains stacking, NIIT and the additional Medicare tax, early-withdrawal penalties, state tax, the Medicare IRMAA tiers, and the ACA subsidy cliff. The marginal rate on the next dollar is measured by re-running the year, so cliffs are visible instead of averaged away.", "It is planning-grade, not filing-grade. No forms, no edge-case completeness, and it says so on the page."]
}, {
q: "Where does the tax law come from?",
a: ["From versioned data files per year and jurisdiction, every figure carrying its citation, its indexing rule, and a confidence tier. The data is content-hashed into each run’s fingerprint, so a projection can always say which law it was computed under. Golden tests pin every cliff against official publications."]
}, {
q: "Which strategies can a plan state?",
a: ["Required minimum distributions, Roth conversion policies with IRMAA, ACA-cliff and 0%-gains guards, substantially equal periodic payments under § 72(t) in all three methods, and Social Security claiming with spousal and survivor benefits. Plans can also establish and fund trusts and gift assets. What each strategy does not model is written down alongside it."]
}, {
q: "Do I have to connect my brokerage accounts?",
a: ["No, and you cannot. There is no aggregator and no credential scraping. You enter holdings into the portfolio tracker, and the planner reads them through one read-only gateway that sees only the portfolios you explicitly link into a plan. The planner never writes to the tracker."]
}, {
q: "Can I check the engine against my own spreadsheet?",
a: ["Yes. Upload a balance-by-year series and the app keeps it, re-runs the diff on every review, and flags divergence beyond a tolerance while stating which lines both sides model. Every projection is reproducible from its own stored inputs."]
}, {
q: "What happens after I adopt a plan?",
a: ["A baseline is frozen: the plan, the assumptions, the balance sheet, the run’s identity, and a sentence you write about why. Once a year a checkpoint measures the year against it, decomposes the drift into its causes, re-runs the alternatives you rejected, and answers in one of three states. “On track, nothing has changed the decision” is meant to be the boring, common one."]
}, {
q: "Is there an AI advisor?",
a: ["An advisor panel exists in the design and is deliberately empty on the page until it is ready. When present, its unsolicited output stays comparative only. The engine, the goals, and the checkpoints work with the advisor off, and the product is built to be walked that way first."]
}, {
q: "Who is this not for?",
a: ["Early accumulators with simple finances, anyone who wants an answer handed to them, traders, and anyone who needs filing-grade tax numbers. Below roughly a couple of million in assets the planning levers are mostly inert, and the complexity does not pay for itself."]
}];
const [open, setOpen] = useState(0);
return