rebyte-skills

Historical data access for the stock-analysis Backtesting pillar. Query the Rebyte Financial Data Service through the Relay Data API with read-only SQL, plus semantic search over historical news when needed for data exploration.

The single financial skill: stock and company analysis plus strategy backtesting, on the Rebyte financial data lake (US equities + China A-shares). Covers price history, technical analysis, company research, news, fundamentals, dividends/splits, insider trading via SEC EDGAR, multi-stock comparison, and NautilusTrader backtesting with a first-class result bundle. Use when the user mentions stock tickers, stock prices, company analysis, investment research, financial data, or wants to backtest a strategy. Triggers: stock symbol (AAPL, TSLA, 000001.SZ), 'stock price', 'analyze stock', 'compare stocks', 'company financials', 'insider trading', 'SEC filing', 'is X a good buy', 'price history', 'backtest', '回测'. Do NOT use for full-blown multi-source research reports (use financial-deep-research instead).

Stock-analysis Backtesting pillar. Run realistic event-driven backtests with NautilusTrader, including historical data acquisition through data/SKILL.md and execution/reporting. Use for strategy simulation, signal evaluation, Sharpe/drawdown/returns, in-sample vs out-of-sample, walk-forward, or parameter comparison. Do NOT use for live trading; the data lake is historical (T+1) only.

Template-only financial analysis workflows for statement analysis, valuation analysis, company research, industry research, and finance-specific output patterns. Use when the user asks for a financial analysis structure, report outline, model rubric, or reusable instructions rather than live data retrieval.

Render TradingView-style interactive price charts (candlesticks, indicator overlays, trade markers, volume/RSI/MACD panes) from OHLCV data using TradingView's open-source Lightweight Charts. Use for any price/market chart - "K线图", "candlestick chart", "price chart", "backtest chart", "plot the strategy", "chart AAPL", "TradingView-style chart", "show entries and exits on the chart". Do NOT use for non-price visualizations (heatmaps, pie, treemaps) - use the echarts skill for those.

Generate music from text (and optionally a reference image) using Google Lyria 3 via Rebyte data API. Two models — `lyria-3` for fast 30-second clips, `lyria-3-pro` for full songs up to ~3 minutes with structural awareness (intro/verse/chorus/bridge). 44.1 kHz stereo MP3 or WAV. Pass an `image` to condition mood/genre on a visual. Use for soundtracks, jingles, intro/outro music, ambient beds, podcast theme music, image-to-music, or any text-to-music task. Triggers include "generate music", "create a song", "make a soundtrack", "compose music", "background music", "jingle", "theme song", "music from prompt", "text to music", "image to music", "music from image", "AI music".

Generate images from text prompts or edit existing images via Rebyte data API. Two backends selectable via the `provider` field — `gemini` (default, Nano Banana 2 / Gemini 3.1 Flash) or `gpt` (OpenAI gpt-image-2). Gemini is best for multi-aspect-ratio output (512px–4K) and fast multi-image edits; gpt-image-2 is best for high-fidelity photorealism and precise text rendering. Supports text-to-image and image-to-image on both. Triggers include "generate image", "create image", "make a picture", "draw", "illustrate", "image of", "picture of", "edit image", "modify image", "enhance image", "style transfer", "nano banana", "gpt image".

Generate stylized presentation slide images using Nano Banana AI. Converts article content to ASCII framework then generates AI images for each slide. Triggers include "nanoppt", "stylized slides", "AI slide images", "presentation images", "Doraemon style slides", "anime presentation". Do NOT use for regular presentations (use slide instead).

Timeline-based motion design — produce short looping or narrative animations. Use when user asks for an animated explainer, intro/outro, kinetic typography, or motion graphic. Output is a self-contained HTML file with CSS/JS keyframes or a Remotion/Lottie composition.

Create a design system or UI kit — visual tokens, typography, color palette, component library, and design principles bundled into a coherent identity. Use when user wants to define a brand, set up styling rules, or build a UI kit before producing artifacts.

Native text and shapes — editable in PowerPoint. Convert an HTML deck artifact to a .pptx where each slide's text, shapes, and tables remain editable downstream. Use when the user needs the deliverable to be opened and modified in PowerPoint or Google Slides. Trade-off: visual fidelity slightly lower than the screenshot variant.

Flat images — pixel-perfect but not editable. Convert an HTML deck artifact to a .pptx where each slide is a high-resolution screenshot. Use when visual fidelity matters more than downstream editability — final pitch decks, hand-off PDFs, archival.

Working app prototype with real interactions — clickable, navigable, with stateful UI. Not a wireframe, not a slide deck. Use when user asks for a prototype, demo app, clickable mockup, or interactive concept. Output is a self-contained HTML/JS app exercising the happy path plus a few key edge cases.

Formal letter — cover letter, resignation, recommendation, outreach, formal memo. Open with purpose, close with ask. Use when the user wants a letter, formal note, or written communication with a clear recipient.

Slide presentation in HTML — narrative-driven deck with one idea per slide, speaker notes, and keyboard navigation. Use when user asks for slides, a deck, a presentation, a pitch, or wants to walk an audience through points. Output is a single HTML file using the project's deck stage component.

Add in-design tweak controls — wrap an artifact's CSS values in a Tweaks block so the user can adjust spacing, color, font size, and other knobs from the Bench without re-prompting. Use when an HTML artifact already exists and user wants to fine-tune visuals interactively.

Single-page summary or brief — concise document hitting one purpose: pitch, exec brief, product overview, or strategy memo. Tight hierarchy, scannable, optimized for a 30-second skim. Use when the user wants a one-pager, brief, exec summary, or single-page document.

CV or résumé — single or two-page document presenting work history, skills, and accomplishments. Numbers beat adjectives. Use when the user asks for a resume, CV, or curriculum vitae.

Print-ready PDF export — convert any HTML artifact (one-pager, resume, letter, deck, report) to a high-quality PDF with proper page breaks, margins, and embedded fonts. Use when the user wants a downloadable, printable, or shareable PDF version of an existing artifact.

Single self-contained file that works offline — produce or convert an artifact into one HTML file with all CSS, JS, fonts, and images inlined or embedded so it works without external assets. Use when the user wants an HTML deliverable they can open offline, email, or archive.

Explore many ideas with wireframes and storyboards — low-fidelity sketches showing layout, hierarchy, and flow without committing to visual design. Use when user is at the divergent stage and wants 3–6 alternative directions or a storyboard of screens. Output emphasizes structure over polish.

Professional slide creating, polish. The slides app at /code/slides/. Create, iterate, and publish HTML or image decks. Each {slug}/ folder is a deck, the code agent is the app's runtime user.

Use this skill for anything PDF — creating professional PDFs (resumes, one-pagers, white papers, letters, portfolios, equity reports, changelogs, custom documents) with the Kami design system; reading or extracting text/tables; merging, splitting, rotating, watermarking, encrypting; filling PDF forms; extracting embedded images; OCR on scanned PDFs. Trigger on any `.pdf` request or phrases like "make me a PDF", "generate a report", "build me a resume", "one-pager", "white paper", "export to PDF", "professional document".

Pair a fast/cheap executor (Haiku, Sonnet, Flash, opencode) with occasional consultations to a high-intelligence advisor (Claude Opus, Gemini 3.1 Pro) when you hit a hard decision point. The advisor runs as a real sub-agent via /api/sub-agent/invoke and returns a plan you continue with. Install when you want near-frontier intelligence at executor-level cost — especially for long-horizon tasks where most turns are mechanical but a few require strong reasoning. Trigger phrases include "advisor", "consult opus", "second opinion from a smarter model", "save tokens with advisor pattern", "stuck on a hard decision".

Convert text to speech audio. Picks from a catalog of OpenAI (gpt-audio) and Gemini voices. Supports style/prosody control — natural language directions for Gemini voices, "instructions" field for OpenAI voices. Use when user wants voiceovers, narration, audio for videos, multi-voice dialogue, expressive or whispered speech.

Add AI/LLM capabilities to deployed applications. Provision an API key for your deployment to access OpenAI-compatible endpoints (GPT, Claude, Gemini) with usage billed to your organization.

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

ReClaw/Zeroclaw-only: create, list, update, and delete scheduled tasks. Only works with the ReClaw (Zeroclaw) agent via messaging channels (Telegram, Slack, etc.). Does NOT work with Claude Code, Codex, Rebyte Code, or other coding agents — those should set schedules in the UI instead.

Access and manage structured data. Query your organization's connected databases, files, and services. Import your own CSV/Parquet/JSON files as queryable SQL tables. Build views that join across sources. All data is queryable via standard SQL.

Gather, clean, analyze, and visualize data from the web or uploaded files. Use when user wants to scrape data, analyze datasets, create charts, find trends, compare metrics, build dashboards, or generate data-driven reports. Triggers include "analyze data", "scrape and analyze", "create dashboard", "find trends", "compare metrics", "data visualization", "statistical analysis", "correlation analysis", "build chart", "data report".

Conduct enterprise-grade research with multi-source synthesis, citation tracking, and verification. Use when user needs comprehensive analysis requiring 10+ sources, verified claims, or comparison of approaches. Triggers include "deep research", "comprehensive analysis", "research report", "compare X vs Y", or "analyze trends". Do NOT use for simple lookups, debugging, or questions answerable with 1-2 searches.

Create and refine visual designs — logos, brand assets, UI mockups, illustrations, interactive artifacts — using AI image generation, design intelligence, and impeccable.style design skills for professional polish. Use when user wants to design a logo, create brand assets, build UI mockups, generate illustrations, or create and refine interactive visual artifacts.

Process, create, convert, and analyze documents — PDFs, Word docs, spreadsheets, and more. Use when user wants to extract data from PDFs, create Word documents, merge or split PDFs, convert between formats, analyze spreadsheet data, or process uploaded files. Triggers include "extract from PDF", "create Word document", "merge PDFs", "convert document", "analyze spreadsheet", "OCR this scan", "extract tables", "create template", "process invoice".

Provision DynamoDB NoSQL database for web apps. Use for key-value storage, document storage, high-performance NoSQL. Triggers include "DynamoDB", "NoSQL", "key-value store", "document database".

Create interactive charts and data visualizations using pyecharts (Python) and Apache ECharts. Use when user needs charts, graphs, or data visualizations rendered as HTML. Triggers include "create chart", "make a graph", "visualize data", "bar chart", "line chart", "pie chart", "scatter plot", "heatmap", "data visualization", "plot this data", "chart this". Do NOT use for static images or matplotlib-style charts.

Embed images, audio, and other assets into a single HTML file using base64 data URIs. Use when building self-contained HTML widgets or pages that need inline media. Triggers include "embed image", "embed audio", "inline asset", "base64 embed", "self-contained HTML", "single-file HTML with media".

Conduct enterprise-grade financial research with multi-source synthesis, regulatory compliance tracking, and verified market analysis. Use when user needs comprehensive financial analysis requiring 10+ sources, verified claims, market comparisons, or investment research. Triggers include "financial research", "market analysis", "investment analysis", "due diligence", "financial deep dive", "compare stocks/funds", or "analyze [company/sector]". Do NOT use for simple stock quotes, basic company lookups, or questions answerable with 1-2 searches.

Build stylish, Typeform-like multi-step forms and surveys using the Rebyte Forms library (Composer API). Outputs standalone HTML files. Triggers include "create a form", "build a survey", "make a questionnaire", "feedback form", "contact form", "signup form", "onboarding flow", "multi-step form", "typeform-style", "data collection form". Do NOT use for simple single-field inputs or backend form processing.

Build forms, surveys, quizzes, and registration pages with multi-step flows, conditional logic, and response collection. Use when user wants to create a form, survey, questionnaire, quiz, signup flow, feedback form, or registration page. Triggers include "create form", "build survey", "make questionnaire", "registration form", "feedback form", "signup flow", "quiz", "typeform", "multi-step form".

Google Workspace workflow for Drive, Docs, Sheets, Gmail, Calendar. Supports two modes - CLI (gws with service account or access token) and Browser (sync Google session via Browser Sync, then automate with Chrome DevTools MCP). Ask the user which method they prefer.