Data analysis, machine learning, LLM apps, and prompt design (9.1K skills)
Set up Acontext so Claude Code or OpenClaw can retain conversations and turn them into reusable skills. Covers CLI installation, login, project selection, plugin configuration, and checks that the configured service is reachable.
Turn financial research requests into Tushare data retrieval, cleaning, comparisons, and concise summaries. Study Chinese stocks, funds, financial statements, market flows, and economic indicators, or export structured files for further analysis.
Analyze casino back-office exports to identify why first-time depositors do not return for a second deposit. The skill groups players by reason, estimates missed deposits, calculates performance metrics, and prepares a recent-player list with suggested CRM actions.
Prepare an AI development phase by selecting a framework, researching official documentation and the application domain, and planning evaluations. The workflow brings these decisions together in an AI-SPEC.md design document.
Verify AI products’ system prompts and tool definitions against dated public archives. Check quoted instructions, identify how records were obtained, and compare products or operating modes using the actual files rather than recalled descriptions.
Integrate AI applications with RouterBase using OpenAI-compatible clients. Plan model selection and fallbacks, meaning alternative models when requests fail, and implement chat, image, video, audio, speech, or embedding workflows with validation steps.
Analyze Mixpanel event counts and trends, compare user groups, and examine funnels, the steps leading to actions such as purchases. The skill also supports searching and updating user profiles through Rube MCP.
Send and search product usage events in PostHog, inspect projects, and create feature flags that control which users see a feature. Use Rube MCP, an agent tool connector, to perform these operations.
Evaluate retrieval and citations in RAG, systems that retrieve documents to support answers, using JSONL test data. Produce repeatable offline metrics, diagnostic reports, and threshold checks for continuous integration without model calls or network access.
Send customer activity and profile data to Segment through Rube MCP. Record events and page views, associate users with organizations, link identities, and submit multiple operations together for customer data collection.
Turn a workflow description into an Envelope AI agent team definition. Design roles and reporting relationships, then generate a structured JSON file with access rules, configuration variables, human approval checkpoints, and recurring schedules as needed.
Retrieve token prices, trading activity, and liquidity pool data from DexPaprika across multiple blockchains. Compare pools, inspect historical price candles, and fetch prices for up to ten tokens together without an API key.
Retrieve GA4 reports for selected dates and metrics, including cross-tabulations and funnels that show where users drop off. List accessible accounts, properties, and key events through Composio’s Google Analytics tools via Rube MCP.
Retrieve cryptocurrency prices, market capitalization, trading volume, historical prices, and exchange information from CoinPaprika. Search for coins or look them up by contract address, with a free tier that requires no API key.
Manage Amplitude product analytics through an AI agent: send event data, review individual activity histories, update user properties, and manage cohorts, groups of users tracked together. It also supports browsing configured event categories.
Establish rules for when an AI agent should recall past context and save lasting decisions, corrections, preferences, and failures. Works with Markdown files or memory services, preserving decision history and surfacing conflicting records.
Extract a focused reference pack from an Obsidian vault for another agent or task. Choose a topic or recent activity, set a token budget to limit length, and receive material with citation information.
Export Obsidian wiki pages and their connections for external analysis or visualization in JSON, GraphML, Neo4j Cypher, SQL, and HTML. Optionally create an OKF Markdown bundle containing page bodies, with project and visibility filters available.
Compare how stocks are discussed on Reddit, X.com, financial news, and Polymarket using the Adanos Finance API. Get attention scores, bullish percentages, activity counts, and trends to identify agreement or differences across sources and stocks.
Query Fintel for short interest, stock borrowing costs, institutional holdings, and insider transactions. Review these alongside analyst forecasts, earnings, and dividends to build a factual market research summary through read-only API access.
Analyze company earnings and revenue forecasts using Yahoo Finance data. Compare analyst consensus, meaning their combined expectations, across periods, track upward and downward revisions, assess forecast ranges, and review how past estimates compared with actual results.
Retrieve stock, crypto, and futures prices, review technical indicators, compare trends across timeframes, and scan exchanges for notable moves. Test trading strategies against historical data through a bundled market data server without a TradingView login.
Retrieve stock prices, historical data, financial statements, dividends, and analyst estimates from Yahoo Finance using Python. Compare multiple stocks, find stocks matching selected criteria, and present the requested figures in readable tables.
Calculate how far an ETF’s market price sits above or below its net asset value, the value of its holdings. Compare funds, screen for large gaps, and investigate sudden moves using holdings data and options-based estimates.