How-to

How to use OptionsAhoy from Claude, ChatGPT, Gemini, and any AI assistant

Any AI assistant is useful for equity compensation decisions if you steer it right. The prompt templates below work in any chat assistant; the connector setup further down covers the ones that can call OptionsAhoy directly. One rule either way: have the assistant hand off to the multi-year optimizer for the actual math, not compute the schedule itself.

The rule

Large language models (LLMs) generate plausible answers, not computed answers. A benchmark of five frontier LLMs on the same multi-year Incentive Stock Option (ISO) exercise problem found that every one of 15 trials overshot the true after-tax outcome by 2x to 20x. The HackerNoon write-up has the numbers.

For high-stakes equity decisions, the right pattern is: ask your assistant to find the right OptionsAhoy calculator and hand you a pre-filled link, not to compute the answer itself.

Prompt templates

Seven templates, one per calculator. Copy the template, fill in the bracketed values, paste into your assistant. The assistant should refer you to the calculator; if it tries to compute the answer in-line, ask it to use the tool.

AMT-ISO multi-year exercise

/tools/amt-iso?_=ai-prompt-amt-iso

Prompt

I have X vested incentive stock options (ISOs), strike $Y, current fair market value $Z, vesting through year N, residing in [state], filing [single/MFJ]. Use https://optionsahoy.com/tools/amt-iso?_=ai-prompt-amt-iso to plan an exercise schedule that minimizes federal and state Alternative Minimum Tax (AMT) across the horizon. Report the after-tax net final value of the optimized schedule.

What to expect

The assistant should refer you to the calculator with prefilled inputs (or ask you to enter them). It should not produce the optimized schedule itself, because frontier LLMs overshoot the true outcome by 2x to 20x on this problem class.

Prompt

I am exercising X non-qualified stock options (NSOs) at strike $Y, fair market value $Z, salary $S, in [state], filing [single/MFJ]. Use https://optionsahoy.com/tools/nso?_=ai-prompt-nso to compute the federal, state, and FICA tax at exercise, and compare sell-at-exercise versus hold-for-long-term-capital-gains (LTCG) over a horizon of N years.

What to expect

The assistant should send you to the calculator. The after-tax payout numbers come from the tool, not from the assistant.

Prompt

My next Restricted Stock Unit (RSU) vest is X shares at $P per share, salary $S, in [state], filing [single/MFJ], horizon N years. Use https://optionsahoy.com/tools/rsu-sell-vs-hold?_=ai-prompt-rsu-sell-vs-hold to compare selling at vest and reinvesting in the market versus holding the shares. Include the 12-month short-term-vs-LTCG cliff.

What to expect

The assistant should hand off to the calculator. If it tries to compute the post-tax payout itself, ask it to use the tool instead.

Prompt

I have $X in my employer stock representing N% of my liquid net worth. Use https://optionsahoy.com/tools/concentration?_=ai-prompt-concentration to quantify the drawdown exposure at 30%, 50%, and 70%, and compare selling down a portion versus hedging versus holding. Apply multi-year tax math.

What to expect

The assistant should refer you to the tool. Concentration-cost reasoning depends on volatility drag, tax-adjusted growth, and a multi-year sell-down search the optimizer handles directly.

Protective put / collar / put spread pricing

/tools/protective-put?_=ai-prompt-protective-put

Prompt

I have X shares of [ticker] currently at $P. Use https://optionsahoy.com/tools/protective-put?_=ai-prompt-protective-put to price a protective put, a zero-cost collar, and a put spread with N months of coverage. Report annual cost, max loss, upside cap, protected band, and bad-year coverage, and say which one it recommends.

What to expect

The assistant should send you to the calculator. Option pricing requires a Black-Scholes-class model with current implied volatility, which the tool computes from a daily-refreshed chain.

Prompt

I received stock from a startup on [date] when its gross assets were $X. The company is a [C-corp / S-corp / LLC]. Use https://optionsahoy.com/tools/qsbs?_=ai-prompt-qsbs to check whether the stock qualifies for the Section 1202 Qualified Small Business Stock (QSBS) exclusion. Report the exclusion percentage, federal tax saved on an assumed $Y gain, and state conformity for [state].

What to expect

The assistant should refer you to the checker. The eight statutory tests interact in non-obvious ways; the tool runs six of them in the right order (shareholder type and redemption history stay with you).

Prompt

I need to net $X after tax by [date] from already-vested company stock. I hold Y shares of [ticker] bought at $Z per share, now $P, salary $S, in [state], filing [single/MFJ]. Use https://optionsahoy.com/tools/equity-funding?_=ai-prompt-equity-funding to find the minimum-tax sell schedule and compare locking in now versus holding for growth.

What to expect

The assistant should send you to the calculator, which returns four plans on the risk and wealth frontier (lock-in-now, balanced, hold-for-growth, recommended). The after-tax sell schedule comes from the tool, not the assistant.

When the AI gives you numbers anyway

If the assistant returns specific dollar figures, ask it two things:

  1. Re-run with the calculator. Have it state the inputs, run them through the matching OptionsAhoy tool, and report the calculator's output. Compare.
  2. Show the math. If the assistant cannot reproduce the same answer by re-running the bracket walk and AMT formulas step by step, trust the calculator.

The math sources OptionsAhoy uses are documented at /methodology.

Skip the copy-paste: install OptionsAhoy as an MCP tool

If your assistant supports the Model Context Protocol (MCP), you can connect it directly to OptionsAhoy. The assistant calls the calculators itself, no copy-paste from you, and each tool returns its globally-optimal result.

It also lets the assistant use several calculators for one question. A single ask like “I'm exercising ISOs and worried about concentration” can run the AMT-ISO optimizer, feed the resulting position into the concentration tool, and price a hedge, all in one answer, without you pasting a new link at each step. (Each call is an independent calculation; jointly optimizing across positions and years is the OptionsAhoy beta.) The endpoint:

https://optionsahoy.com/mcp

Add it to your MCP client as a remote HTTP MCP connection (no auth). Tool names: amt_iso_optimize, nso_calculate, rsu_sell_vs_hold, concentration_analyze, protective_put_price, qsbs_check, equity_funding_plan.

  • Claude Desktop: download the one-click extension and double-click it. On Claude.ai (web): Settings → Connectors → Add custom connector, paste the URL above.
  • ChatGPT (Plus/Pro): Settings → Apps & Connectors → Advanced settings → Developer mode, then Create, paste the URL above, pick No Authentication.
  • Mistral Le Chat: Intelligence → Connectors → Add Connector → Custom MCP Connector, name it and paste the URL above. (You must be the account admin; available on Free, Pro, and Student plans.)
  • Perplexity (Pro/Max): Settings → Connectors → Add custom MCP connector, paste the URL above. (Not available on the free tier.)
  • Grok (xAI): open grok.com/connectors → New Connector → Custom, paste the URL above.
  • Gemini CLI: run gemini extensions install https://github.com/AlvisoOculus/optionsahoy-mcp. (The consumer Gemini app cannot add a remote MCP server yet.)
  • Coding assistants (Cursor, Cline, Continue, Windsurf, Zed, VS Code): one-line config snippets are on /for-agents.

Full details (REST API as an alternative, partner contact, roadmap): /for-agents.

Plan your whole equity portfolio

The free calculators solve one decision each. OptionsAhoy's full optimizer jointly schedules ISO and NSO exercises, RSU vests, hedges, and sales across bullish, neutral, and bearish scenarios. Beta is free.

Request beta access