AI agent skill for options trade analysis — pick the right structure for your view, compute Max P/L, break-evens, Greeks, IV-rank context, PoP-vs-EV & risk management. Works with Claude Code, Codex, Cursor. By Viprasol Tech.
SaferSkills independently audited options-strategy-analyzer (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A disciplined, repeatable methodology for vetting an options trade before you click buy — or for picking the right structure given a market view. Most losing options trades aren't bad analysis of the underlying; they are a structure mismatch (long premium in a 70 IV-rank market, short premium in a 20 IV-rank market), a size mismatch (a 5% account-max-loss trade dressed up as a "small" position), or an assignment / liquidity surprise the trader never priced in.
The core skill is naming the structure precisely, computing the closed-form math (max profit, max loss, break-even, R:R), reporting the net Greeks (Δ, Γ, Θ, ν) and what they cost you per day, contextualizing with IV rank, estimating probability of profit honestly (with the "PoP ≠ EV" caveat), and enforcing the risk-management & operational checks that separate profitable short-premium operators from blown accounts.
Honest scope & limits — read this first.
>
I do not have a live options chain or quotes. I reason over what you provide: underlying ticker, current price, account size, strikes, expirations, implied volatilities, mids / credits / debits, and any existing legs. Without numbers, output is qualitative (structure selection + risk framing only). The math below is at expiration unless an intraday Greeks-based P&L is explicitly requested.
>
This is not investment advice. Options carry the risk of losing 100% of premium paid, and undefined-risk positions (naked short calls, short strangles, short straddles) can lose more than the cash deposited. Past win rates do not predict future results — a strategy with a historical 70% win rate can take a tail loss on its next trade. Use at your own risk and consult a qualified professional.
Activate when the user:
debits, IVs) and asks "is this a good trade?" or "what are my max profit / max loss / break-evens?".
volatility view (IV high / low / neutral) and wants a strategy recommendation.
volatility, IV rank, IV percentile, debit, credit, spread, vertical, bull / bear call / put spread, iron condor, iron butterfly, strangle, straddle, calendar, diagonal, covered call, cash-secured put, short premium, long premium, theta, gamma, vega, delta, PoP, probability of profit*.
or managing an existing position (closing at a profit target, rolling, taking a tested side off).
You cannot fetch the chain. This skill reasons over what the user gives you. If you need strikes, IVs, mids, account size, or the user's view, ask for them (Step 1) — never invent prices, deltas, or volatilities.
Establish exactly what you're analyzing before saying anything about edge.
futures option), spot price, and lot size if not 100 (e.g. mini index options are 10×).
enforce position sizing.
(intraday / days / weeks / a cycle).
neutral. Ideally with IV rank and IV percentile numbers.
index rebalance — anything that can spike IV or jump the underlying inside your expiration window.
monthly (~30–45 DTE — the credit-spread / iron-condor sweet spot), long-dated (LEAPS — vega-heavy, low theta).
position, or managing to an exit?
Do not proceed to a verdict until you know what the underlying is, what the view is, and what the structure is (or which structures you're choosing between).
The decision framework is directional view × volatility view, then defined vs undefined risk. Mark every candidate as defined or undefined, and give a one-line best when / worst when before picking.
| Long Vol (buy premium — long Γ, long ν, short Θ) | Short Vol (sell premium — short Γ, short ν, long Θ) | |
|---|---|---|
| Bullish | Long call · Bull-call debit spread | Cash-secured put · Bull-put credit spread · Covered call |
| Bearish | Long put · Bear-put debit spread | Bear-call credit spread |
| Neutral | Long straddle · Long strangle · Calendar / Diagonal (long-vol leg) | Short strangle · Iron condor · Iron butterfly · Calendar (short-vol view) |
loss is bounded and computable at entry. The width − credit (or the debit) is your worst case.
straddle): margin-intensive, real assignment risk, and loss can exceed the credit collected — for a short call the loss is theoretically unlimited on the upside.
Quick guide for the most-common picks:
when IV rank > 50 and you have a clear support level under the short put. Worst when* IV is low (credits don't pay you for the downside).
rank > 50 and the underlying is rangebound with no event in window. Worst when IV is rising into a binary event you're holding through.
cheap premium. Best when IV rank < 30 and a catalyst is coming. Worst when IV rank > 60 — you are overpaying for vol that will compress.
hold; income against a cap. Best when IV rank > 40 and the strike is at your willing-to-sell price. Worst when the stock is set to run hard (you cap the upside) or to crash (the small premium doesn't hedge you).
risk* and margin requirement. Best when IV rank > 60 in a liquid, rangebound underlying with no events. Worst when* a gap blows past your short call or short put — losses can be many multiples of the credit.
For the chosen structure, output a Math table: net debit / credit, max profit, max loss, break-even(s), R:R (max profit ÷ max loss), and width for spreads. If a number is missing, compute symbolically and ask for the missing input. Always show the formula, then plug in numbers.
Sell put at strike K_short, buy put at strike K_long, with K_long < K_short. Per-share credit C. Width W = K_short − K_long.
Max profit = C
Max loss = W − C
Break-even = K_short − C
R:R = C / (W − C)(P&L per contract = the per-share figure × 100 for standard US equity options.)
Sell call at K_short, buy call at K_long, with K_long > K_short. Per-share credit C. Width W = K_long − K_short.
Max profit = C
Max loss = W − C
Break-even = K_short + C
R:R = C / (W − C)Short put K_sp, long put K_lp, short call K_sc, long call K_lc — with K_lp < K_sp < K_sc < K_lc. Net credit C. Wing width W = K_sp − K_lp = K_lc − K_sc (assume equal wings).
Max profit = C (when K_sp ≤ S ≤ K_sc at expiry)
Max loss = W − C (when S ≤ K_lp or S ≥ K_lc)
Break-evens = K_sp − C and K_sc + C
R:R = C / (W − C)(Unequal wings — the "broken-wing" condor — change max loss to max(put_width, call_width) − C. Flag this if the user constructs asymmetric wings.)
Buy call at strike K for per-share debit D.
Max profit = unlimited (theoretical)
Max loss = D
Break-even = K + DBuy put at strike K for per-share debit D.
Max profit = K − D (if S = 0 at expiry)
Max loss = D
Break-even = K − DLong 100 shares of underlying at cost basis S₀, short one call at strike K for per-share premium P.
Max profit = (K − S₀) + P (if S ≥ K at expiry — called away)
Max loss ≈ S₀ − P (large; the downside of holding the stock,
buffered only by the small premium)
Break-even = S₀ − PThe covered call is a bullish-to-neutral, capped-upside structure. The "max loss" line is intentionally large — owning stock is the dominant risk; the premium is a small income buffer, not a hedge.
Sell put at K_sp, sell call at K_sc, with K_sp < K_sc. Per-share net credit C (sum of the two premia collected).
Max profit = C (when K_sp ≤ S ≤ K_sc at expiry)
Max loss = theoretically unbounded on the call side
(and very large on the put side, down to S = 0)
Break-evens = K_sp − C and K_sc + C
Margin = broker formula (~20% × underlying for short call side,
minus OTM amount; check your broker)A short strangle's max profit line is the entire compensation; the max loss line is the reason it requires margin, position-size discipline, and a plan for a managed exit before a tail event.
R:R note. For credit spreads, R:R is mechanically poor (you are paid small to take a defined-but-larger risk). The right framing is not "is R:R > 1?" — it's whether the probability of profit and your management plan together make the expected value positive after fees.
For each leg, list Δ (sign and magnitude). Then report the net position Greeks with sign and rough magnitude per 1-contract structure.
| Greek | Means | Sign you want depends on view |
|---|---|---|
| Δ | ∂Price / ∂Underlying | Bullish → Δ > 0; bearish → Δ < 0; neutral target Δ ≈ 0 |
| Γ | ∂Δ / ∂Underlying (convexity) | Long-premium → Γ > 0; short-premium → Γ < 0 |
| Θ | ∂Price / ∂time (per calendar day) | Short-premium → Θ > 0 (decay works for you); long-premium → Θ < 0 |
| ν | ∂Price / ∂IV (per 1 vol-point) | Short-vol → ν < 0 (IV drop helps); long-vol → ν > 0 (IV rise helps) |
Conventions:
implied distribution — useful for strike selection (e.g. a 16-delta short strike ≈ a 1σ move).
convexity (Γ) and vol exposure (ν) and pay time (Θ). Short options collect Θ and pay both Γ and ν. A "free" Greek combination doesn't exist.
premium, and the reason monthly cycles are typically managed by 21 DTE.
Report Θ as a per-day number ("≈ +$X / day at entry, all else equal") and ν as a per-1-vol-point number ("≈ −$Y per +1 IV point"). If the user hasn't given Greeks, state the signs and ask for magnitudes only if needed.
If IV rank or IV percentile is provided, apply the volatility heuristics — IV is the most under-used input in retail options decisions.
0..100.
Rules of thumb (from the research):
credits; vol tends to mean-revert lower). Credit spreads, iron condors, short strangles get paid more.
options; vol may expand). Long calls / puts / debit spreads have better entry cost.
directional / structural merit alone.
Two refinements when relevant:
higher than OTM call IV ("put skew"). This matters for strangles and iron condors — your put-side credit is naturally richer than the call-side credit at equal deltas, and your put-side risk is correspondingly larger.
is normal. Backwardation (front > back) usually signals an event in the front cycle (earnings, macro) and is a warning for calendars and diagonals — the short-vol leg can blow up at the event.
Estimate probability of profit (PoP) honestly, and state the caveat in the same breath.
For a short-premium trade (one short option):
PoP ≈ 1 − |Δ_short|(Delta-as-probability heuristic. A 25-delta short strike → ≈ 75% PoP.)
For a defined-risk credit spread:
PoP ≈ 1 − (Max loss / Width) = 1 − (W − C) / W = C / W(The premium-implied probability that price stays past the short strike at expiry. Both estimates should agree to within a few points; if they don't, report both and note the divergence.)
For an iron condor (two-sided short premium), PoP is approximately:
PoP ≈ 1 − (|Δ_short_put| + |Δ_short_call|)(Both wings need to stay OTM; the failure probabilities add.)
PoP ≠ EV — say this every time. These estimates are the option chain's premium-implied probabilities. If implied volatility is fair, the expected value of a short-premium trade is ≈ 0 before fees and slippage. A 90% PoP trade with a 9× downside is not edge — it is risk transfer, and if IV is honestly priced you are not being overpaid for it. Reported historical win rates (e.g. "65–70% on credit spreads") describe a win rate, not an edge: a single tail loss can eat many winners.
What this means in practice:
to judge whether you are being paid for the risk.
premium-rich because the market structurally fears down-gaps; sellers can be compensated for that risk, but they own that risk.
These rules separate operators from blow-ups. Run all of them every trade.
position_size_% = max_loss_per_contract × contracts / account_size.
desks size at ≤ 1–2% per defined-risk trade, and well under 5% even for undefined-risk structures.
Kelly assumes your edge estimate is exactly right, and a wrong edge estimate produces dramatic drawdown.
half its credit has already realized most of the easy theta with most of the gamma risk still to come. Closing early raises annualized return on capital.
outweighs the remaining theta. Roll the position to the next cycle, take it off, or take the untested side off and let the tested side run with a defined hard stop.
wing closer is often a fool's game — it adds risk to a side that wasn't in trouble.
all of their gain in days as Θ accelerates near expiry. Scale out into strength.
is a worsening situation; do not turn it into a hope-trade.
risk — the holder of the long call can exercise the day before the ex-date to capture the dividend. If your short call has less extrinsic value than the dividend, expect to be assigned.
any day. Index options (SPX, NDX, RUT) are European — exercise only at expiry — and cash-settled, removing assignment risk at the cost of PM-settlement (SPX AM vs PM) timing nuance.
strike at expiry can leave you with an unexpected stock position Monday.
rule of thumb, bid-ask spread > 10% of the mid is illiquid for retail. A $0.20-wide market on a $1.00 credit gives up ~$20 of edge per contract at entry and another $20 at exit — that's your week's profit.
contracts is not. Stale OI ≠ live liquidity — check today's bid-ask too.
the broker's margin requirement (~20% of underlying notional minus OTM amount, plus the premium received — varies by broker) is your real capital commitment, not the credit. Size against margin, not against max-loss-of-credit.
Any one of these flips the trade toward 🟠 / ⛔.
loss — the trade is structurally underpaid.
is likely to compress on top of any directional move you need.
for finite credit. Never structural; always tactical.
explicit volatility plan. Implied vol rises into the event and crushes after — short-premium positions held through the event can earn the vol crush, but only if you've sized for the gap and decided in advance whether you'll hold or close.
theta), or too long for the catalyst (capital-inefficient; vol decay hurts the long-premium side).
failures, not analysis failures.
You will pay 2× spread to round-trip.
amazing on the win column and ugly on the math — the single loss eats 9 wins. Don't conflate PoP with edge.
For cash-secured puts / covered calls / spreads that can pin to a stock position, ensure the assigned position is one you'd actually hold or short.
Lead with a verdict banner, then the math, Greeks, IV context, PoP, management plan, and red flags. Close with the disclaimer.
defined risk within size limits, no red flags. State the one reason it works.
inputs are weak (IV regime borderline, R:R thin, marginal liquidity). State the caveats — fix or accept.
IV regime, oversized, event in window without a plan). State the issue and the better alternative.
thesis, IVR < 20 short premium with rich downside, > 5% account max loss, illiquid). State the reason in one line.
| Metric | Formula | Value |
|---|---|---|
| Width (if spread) | K_short − K_long (or wing width) | … |
| Max profit | … | … |
| Max loss | … | … |
| Break-even(s) | … | … |
| R:R | max_profit / max_loss | … |
| Greek | Position | What it costs / pays |
|---|---|---|
| Δ | + / − / ≈ 0 | directional exposure per $1 underlying move |
| Γ | + / − | convexity sign |
| Θ | + / − | ≈ $X per calendar day at entry |
| ν | + / − | ≈ $Y per +1 IV point |
before fees."**
max_loss × contracts / account_size = …%. PASS /REDUCE TO N CONTRACTS.
scale out (long premium).
early-exercise risk.
Bulleted; if none, say so.
Educational analysis — not investment advice, not a recommendation to trade, and not a guarantee of outcome. Options can lose 100% of premium paid, and undefined-risk positions can lose more than the cash deposited. Past win rates do not predict future results, and the probability estimates above are the option chain's premium-implied probabilities, not measures of edge. This analysis reasons only over the information you provided; it cannot see live quotes, the live chain, or live IV. Verify all numbers in your broker before entering. Consult a qualified, licensed financial professional for personal investment decisions.
options-strategy claim (look-ahead bias, overfit, survivorship, slippage, risk-of-ruin).
sizing, Kelly fraction, drawdown control, and correlation across positions.
Not affiliated with or endorsed by Anthropic.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.