Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Crypto Portfolio Risk Guards

    1

    Crypto Portfolio Risk Guards gives your agent a measured, repeatable risk check of a crypto portfolio — not a trade signal.

    $8

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Crypto Portfolio Risk Guards

    Crypto Portfolio Risk Guards

    Example session with this skill installed

    Check the risk of my crypto portfolio (BTC, ETH, SOL, LINK) against my guard limits and report in Spanish.

    • Read your context and instructions
    • Compiled the crypto portfolio risk

    Live local demo, 10 November 2026, 23:46 Europe/Madrid. The portfolio is an EXAMPLE with illustrative holdings (0.07 BTC, 2 ETH, 40 SOL, 300 LINK; stops on BTC, ETH and LINK) priced from public Binance daily klines labeled /USD (not Twelve Data). Window: 10 Nov 2025 to 9 Nov 2026, 364 daily returns, all bars closed.

    Value 16,214.77 USD. Weights: BTC 32.5%, ETH 26.3%, SOL 21.4%, LINK 19.7%. HHI 0.260, effective N 3.85, top-3 80.3%. Pairwise correlations ran from 0.82 to 0.91 (average 0.87), so at a 0.80 threshold all four coins formed one cluster holding 100% of the portfolio against a 65% limit: BREACH.

    Portfolio annualised volatility 53.6%, beta to BTC 1.16. One-day historical VaR 95 was 4.46% (723 USD) and CVaR 95 6.03% (978 USD); VaR 99 was 6.86% (1,112 USD) and CVaR 99 9.41% (1,526 USD). Max drawdown with today's weights was 53.0%, and the portfolio sat 36.4% below its window peak against a 25% limit: BREACH.

    Guards: 8 PASS, 2 WARN, 5 BREACH. BTC weight WARN (32.5% vs a 40% limit, warning from 32%); the stop-risk guard flagged ETH (2.90% of portfolio to its stop), SOL (2.18% using a 2-ATR distance, no stop set) and LINK (3.08%) against a 2% limit. Risk band: HIGH (11 points).

    The report lists the source, window, and data notes, and it does not tell you to buy or sell. You decide what to do.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Crypto Portfolio Risk Guards gives your agent a measured, repeatable risk check of a crypto portfolio — not a trade signal.

    You tell the agent what you hold: symbol, quantity, optional cost basis, optional stop. The agent pulls daily candles (on Agensi, typically through Twelve Data after you approve the tool quote) or uses candles you paste. A standard-library Python script then runs offline: no network, no API keys, no exchange account, no IP allow-listing. Nothing here can touch your funds.

    You get:

    • Portfolio value, weights, and unrealized PnL against your cost basis. Stablecoin balances are treated as cash.
    • Concentration: HHI, effective number of holdings, top-1 and top-3 weight.
    • A correlation matrix and correlation clusters, with the combined weight of each cluster, so you can see when four coins behave like one.
    • Annualised volatility, 1-day historical VaR and CVaR at 95% and 99% (in % and in currency), max and current drawdown, and beta to BTC.
    • ATR(14) risk-budget context per position: how much of the portfolio a 2-ATR move would cost, and the weight that fits your per-position budget. It is a yardstick, not an order size.
    • Guards you configure: max weight per asset, max correlated cluster, max drawdown, VaR limit, stop distance and risk-to-stop. Each returns PASS, WARN, or BREACH, and the report ends with a risk band (LOW, MODERATE, ELEVATED, HIGH).
    • A report written in your language, with window dates, data source, data notes, and a clear disclaimer.

    Built to be checked: the math is in one readable script under 20 KB, deterministic, and covered by tests that recompute VaR, CVaR, correlation, drawdown, and beta independently with numpy and pandas.

    Requirements: Python 3.10+ (standard library only) and at least 60 overlapping daily returns across your assets (about 365 recommended). Long-only spot holdings. This skill is read-only: it does not place or change orders, and it is not financial advice.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Trading & Investing.

    Analysis only, no orders, no rebalancing advice. Static-weight historical VaR; needs ~1 year of daily closed candles per asset. Long-only spot. Not financial advice.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 6 days ago

    Software
    Python

    Permissions

    Terminal / Shell
    Read Files
    Listed6 days ago

    What's inside

    Frequently Asked Questions

    While most portfolio trackers only show current balances and PnL, this skill analyzes historical volatility and asset correlations to predict potential losses (VaR) and identify 'hidden' concentration risks where assets move in lockstep.

    The skill is designed for any AI agent or platform that supports Python execution and Agensi MCP tools. It functions as a read-only analysis tool, meaning it can process data provided directly in a chat or fetched via integrated market data APIs.

    To ensure statistical significance, the engine requires at least 60 overlapping daily price candles for each asset in your portfolio. This allows the Python engine to accurately calculate correlation matrices and historical volatility without skewing the results.

    No, the engine does not require API keys for your exchange or wallet. It operates on a read-only basis using the symbols and quantities you provide, then fetches public market data to perform the risk calculations locally and securely.

    You receive the full Python-based risk engine, the integration logic for market data retrieval, and a set of pre-configured risk guardrails (VaR, CVaR, and ATR) that can be triggered by your AI agent during a portfolio audit.

    The current version is optimized for spot holdings only. It does not currently support the complex margin requirements, funding rates, or non-linear delta associated with leveraged positions, shorts, or options.

    New in Trading & Investing