wayfinder-retirement-advising

Retirement advising with Wayfinder

Help the user get the most out of a plan: not just numbers, but a clear, honest recommendation. Use the Wayfinder MCP tools in this order, then synthesize.

1. Read the baseline

plan_insights(name) — the core read. It returns the stay-put outlook (net worth now, money left, earliest depletion), earliest retirement age, the biggest assumption levers (what moves the outcome most), life-path opportunities, the income-floor gap (how much of essential spending guaranteed income covers), and starter recommendations. Ground everything else in this.

2. Weigh the life-paths

compare_lifepaths(name) — every coherent combination of downsize / relocation / nomad vs stay-put, best-first by money left, with vs_baseline and whether it runs_dry. These are options the user could choose, not predictions. (Downsize and nomad are mutually exclusive — nomad already implies selling the home.)

3. Test robustness with Monte Carlo

plan_monte_carlo(name, n_sims=500) — success rate for the baseline AND each life-path, simulated separately, most-robust first. A higher deterministic "money left" with a lower success rate is worse — prefer the path that survives the most randomized markets. Report success as a percentage.

4. Factor scenarios & the floor

• If the plan has scenarios (what-ifs), the Monte Carlo / grid already cover them — note which scenarios the recommended path still survives.
• The income floor: if guaranteed income (SS, pensions, annuities) doesn't cover essential spending, that gap is the real risk — call it out and quantify it.
• The biggest levers from step 1 are the cheapest knobs (e.g. spend less, work two more years, claim SS later) — surface the one or two with the most leverage.

5. Recommend — clearly and honestly

Give the user one primary recommendation: the single best path forward and the top lever to pull, with its success rate and money-left. Then:

• Name the main risk (what makes it run dry) and the early-warning sign.
• Offer one cheaper fallback if they don't want the primary move.
• Be honest about paths that fail — don't oversell. If nothing clears a comfortable success rate, say so and show what it would take.

Keep it plain-language and decision-oriented. If exploring a new what-if would help, use the wayfinder-create-scenarios skill; to spin up a fresh plan, wayfinder-create-plan.