
Use when tasks involve reading, creating, or reviewing PDF files where rendering and layout matter; prefer visual checks by rendering pages (Poppler) and use Python tools such as `reportlab`, `pdfplumber`, and `pypdf` for generation and extraction.
Review code against specs for quality, correctness, modularity, and adherence to ward conventions.
Run code against live data and evaluate output quality with domain expertise. Spot-check values and verify completeness.
Build event and catalyst intelligence directly from yfinance news, earnings calendars, analyst actions, and price reaction windows. Use when an agent needs to explain price moves, identify upcoming catalysts, or map sentiment shifts to market behavior.
Write clean, modular, reusable code in wards. Use write_file and edit_file tools for all file operations.
Search the web, news, and images using DuckDuckGo, and convert any webpage into clean readable markdown. Use this skill whenever the agent needs to search the internet, look up current information, find news articles, search for images, or read/extract content from a URL. Trigger when the user says "search for", "look up", "find", "what's the latest on", "read this article", "get the content from", "summarize this page", or any request requiring web information the agent doesn't already know. Also trigger for research tasks, fact-checking, gathering context for decisions, reading Reddit posts/threads, or extracting article text from any URL. This is the agent's primary tool for accessing web information. Do NOT use for Polymarket-specific queries (use polymarket-trader) or for financial market data (use yfinance/ML skills).
Transform any AI-generated or data-dump HTML report into a premium, visually distinctive, dark-themed report with professional typography, animated cards, gauges, data bars, and a coherent design system. Use this skill whenever a user wants to: enhance or beautify an existing HTML report, create a polished research report from raw data, produce a professional dashboard-style document, or make any data-heavy output look publication-ready. Triggers include: "enhance this report", "make this look professional", "beautify this HTML", "create a report for X", "turn this data into a report", "design a dashboard for", "deep research report", or any upload of a plain/ugly HTML report asking for improvements. Apply to any domain: finance, analytics, research, health, engineering, project status, sales, HR, product metrics, or anything else.
Build company and asset fundamental dossiers directly from yfinance objects such as info, financial statements, earnings, and analyst metadata. Use when an agent needs valuation, growth, profitability, leverage, or earnings-quality analysis for equities.
Analyze options chains directly from yfinance to measure implied volatility structure, open-interest positioning, and directional options sentiment. Use when an agent needs expiry-by-expiry options diagnostics, put-call ratios, max pain estimates, or volatility regime checks.
Perform portfolio construction and risk diagnostics directly from yfinance multi-asset return series. Use when an agent needs correlation analysis, drawdown and VaR metrics, stress scenarios, or weight optimization for stock, ETF, and crypto portfolios.
Helper tools and instructions for creating new agents
Lightpanda browser, drop-in replacement for Chrome and Openclaw default browser - faster and lighter for tasks without graphical rendering like data retrieval. Use it via MCP server, CLI fetch, or CDP with Playwright/Puppeteer.
Visual intelligence skill — use when asked to analyze images, screenshots, diagrams, charts, PDFs with visuals, or any task requiring sight. Extracts structured insights from visual content using the multimodal_analyze tool. Works even when the current agent runs on a text-only model.
Build technical indicator, trend, and momentum signals directly from yfinance price data. Use when an agent needs RSI, MACD, EMA, Bollinger, or ATR workflows, multi-timeframe confirmation, signal scoring, or chart-ready technical outputs.
Collect, normalize, and validate market data directly from yfinance for stocks, ETFs, indices, forex, crypto, and futures. Use when an agent needs reliable OHLCV downloads, multi-ticker batching, timezone-safe alignment, corporate action handling, or clean datasets for downstream analysis.