How to use the Crypto Derivatives Strategies agent skill

What the skill makes the agent do

Agent skills are folders containing a SKILL.md file with instructions the agent reads when a request matches the skill's description. This one holds a playbook on three areas: perpetual funding-rate arbitrage, futures term-structure trading, and options strategies such as volatility smile and Greeks analysis.

Once installed, the agent can explain these concepts, interpret funding-rate levels as signals, compute basis and annualized basis from prices you supply, and walk through options structures such as short straddles, protective puts and iron butterflies with suggested parameters. The skill is static text: it does not fetch prices and does not place trades.

• Explain perpetual funding-rate mechanics and positive or reverse carry arbitrage
• Classify funding-rate levels into market sentiment and strategy signals
• Compute basis and annualized basis from spot and futures prices
• Describe contango and backwardation and the related calendar-spread ideas
• Explain Greeks (Delta, Gamma, Theta, Vega, Rho) and the 25Δ risk reversal
• Walk through short straddle, protective put and iron butterfly structures

Install the skill from the ZIP file

Download the ZIP file from this directory page. Unzip it so that the crypto-derivatives folder, containing SKILL.md, sits inside your skills directory. For a personal setup, use ~/.claude/skills/. For a single project, use the .claude/skills/ folder inside that project.

If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills.

• Personal: ~/.claude/skills/crypto-derivatives/
• Project: .claude/skills/crypto-derivatives/ inside your project
• Claude apps: Settings → Capabilities → Skills → upload

Trigger the skill with natural requests

You do not need a special command. Describe what you want in plain language that matches the skill's topics: funding rates, basis, term structure, volatility smile, Greeks or options structures. The agent should pick up the skill when your request fits its description.

Name the exchange or asset when it matters, since the playbook focuses on BTC and ETH on OKX and Deribit. Supply the numbers you want processed, because the skill cannot look them up.

• "Using a funding rate of 0.02% per 8 hours, what is the annualized return?"
• "Explain the difference between contango and backwardation in BTC futures."
• "Walk me through a short straddle on BTC and its maximum loss profile."
• "Compute annualized basis for the BTC quarterly future at 107000 with spot at 105000."

Use the basis and annualization formulas

The skill includes a Python function, term_structure_spread, that takes a spot price and a dictionary of futures prices by expiry, then returns the basis, annualized basis and days to expiry for each contract. It uses pandas (pd), so that library must be imported in your own code. Treat it as a starting point for a crypto analysis tool, not a finished product.

Funding is annualized in the source as funding rate × 3 × 365, because OKX settles every 8 hours.

Understand the risk controls and limits

The file lists the main risks of carry arbitrage: insufficient margin leading to liquidation, funding reversals, basis moves that cause floating losses, and exchange risk such as withdrawal limits and downtime. It also suggests parameters such as leverage no more than 3x and a margin ratio above 50%.

These are suggestions in the playbook, not validated performance claims. Check them against your own exchange rules and risk tolerance before relying on them.

• Leverage: no more than 3x, per the source's suggestion
• Margin ratio: keep above 50%
• Single-coin allocation: under 30%
• Stop-loss: close when floating loss exceeds expected return over 3 months

Safety notes: scripts and network

The skill runs no scripts and uses no network access, and it requires nothing to be installed beyond the file itself. Its only file is SKILL.md. The Python snippet in the text is an example for you to run in your own environment; the skill does not execute it.

Because the skill makes no network calls, it will not show current prices, funding rates or open interest. Any figure you act on needs to come from your exchange or a trusted data source.

• No scripts are run by the skill
• No network access is used
• Files: SKILL.md only
• Licence: MIT