xy121718/baoyu-skills
6 skills6 installs564 starsGitHub
Install
npx skills add https://github.com/xy121718/baoyu-skillsSkills in this repo
1Baoyu Article Illustratorbaoyu-article-illustrator is an agent skill that turns article copy into cartoon-style infographic illustrations for solo creators shipping blogs, newsletters, and growth content. It enforces clear image specs—landscape 16:9, hand-drawn-only visuals, concise keyword-forward layout, and typography that matches the article language—so outputs stay on-brand and scannable. A blocking first-time setup flow collects illustrator preferences in EXTEND.md before the agent asks for references or drafts, which prevents half-configured runs. Use it when you have finalized or near-final prose and need a consistent visual hero or inline graphic without hiring a designer for every post. It pairs with content workflows in Grow and Launch when you repurpose the same article across channels.1installs2Baoyu Cover Imagebaoyu-cover-image is an agent skill that turns your post topic and keywords into a structured cover-image brief, then applies built-in auto-selection rules for visual type, color palette, and rendering style. Solo and indie builders use it when they need consistent visuals for launches, architecture explainers, quotes, stories, or minimal product messaging but do not want to micromanage design tokens on every draft. The skill reads signal tables in SKILL.md—product launch language steers toward hero and vivid palettes, while API and system content skews conceptual with cool tones and flat-vector rendering. It fits creators shipping on WeChat-style channels, infographics, or polished SaaS marketing pages. You invoke it during content production or right before distribution so every piece has a deliberate cover without hiring a designer for routine posts. Confidence is moderate where the public readme is truncated, but the documented rule grids are explicit enough for reliable tagging and discovery.1installs3Baoyu Danger Gemini WebBaoyu-danger-gemini-web is an advanced integration skill packaging an unofficial browser-session Gemini web client for agents that need multimodal output without provisioning official API credentials. The embedded implementation covers access-token and cookie lifecycle, file upload helpers, candidate parsing, gem plugins, and defensive error handling for rate limits, temporary blocks, and model invalidation. Solo builders typically reach for it when prototyping generative features, automating content drafts, or experimenting with image generation inside Claude Code–driven pipelines. Because it depends on fragile web auth semantics, it demands careful secret handling and is a poor fit for production compliance-heavy deployments; treat it as a build-time bridge until you migrate to supported APIs.1installs4Baoyu Image Genbaoyu-image-gen’s first-time setup skill walks solo builders through mandatory configuration of default image generation provider and model before any image is produced. It targets agents using the baoyu skills stack who lack EXTEND.md or have a null default_model for the active provider. The workflow branches on whether configuration is missing entirely versus only the model slot, always collecting answers through AskUserQuestion in one batch rather than assuming defaults. After confirmation it writes or updates EXTEND.md and hands off to generation. That gate prevents wrong-provider spend, broken auth against a transit API, and opaque failures when the main skill runs. Best for indie builders adopting baoyu-image-gen for the first time or rotating models; skip when EXTEND.md already has a valid default_model for your provider.1installs5Baoyu Infographicbaoyu-infographic is a journey-wide agent skill that runs a deep content-analysis pass before any infographic is designed. Solo builders and indie marketers invoke it whenever dense material—launch posts, API guides, competitive research, or onboarding docs—must become a visual that viewers can scan without losing accuracy. The framework treats the analyst as an instructional designer: read the whole source, set learning objectives, classify content type (timeline, process, comparison, and related patterns), then recommend layout and style combinations such as linear-progression with craft-handmade or ikea-manual aesthetics for tutorials. It explicitly forbids paraphrasing or diluting facts, which makes it suitable for ship-phase explainers and grow-phase content equally. Builders use it to avoid pretty but wrong graphics and to align structure with how people actually retain steps and comparisons.1installs6Baoyu Xhs ImagesBaoyu XHS Images is an agent skill for solo creators and indie marketers who publish on Xiaohongshu and want repeatable, on-brand image carousels instead of one-off chat generations. When EXTEND.md is missing, the skill forces a gated setup: watermark choice, default style, and related preferences are collected through AskUserQuestion with explicit options—defaults are not applied without user confirmation. That blocking step exists so later content analysis and image generation honor stable brand rules. After setup, the workflow continues into article-driven visual production aligned with baoyu-xhs-images conventions. It suits builders who treat social content as a growth loop and need procedural guardrails so agents do not rush past configuration. Skip it if you only need generic social graphics without XHS-specific preference files.1installs