zc277584121/marketing-skills
13 skills11k installsGitHub
Install
npx skills add https://github.com/zc277584121/marketing-skillsSkills in this repo
1Browser ScreenshotBrowser Screenshot is an agent skill for solo builders and small content teams who need trustworthy visuals without hand-cropping in a design tool. It assumes agent-browser is installed and Chrome remote debugging is enabled, then runs a deliberate pipeline: find the best URL or post with text tools first, open the page only for capture, select the DOM region that matches the narrative, and export a tight crop suitable for articles and social posts. The skill encodes a hard quality gate—full-screen captures are explicitly forbidden because nav bars, sidebars, and ads dilute the story. That makes it especially useful when you are writing case studies, comparison posts, or launch threads where a single tweet or chart must read clearly at thumbnail size. It is not a general browsing assistant; research stays in fetch/search tools until the target is validated.936installs2Remove Ai StyleRemove AI Style is an agent skill for solo builders who draft marketing and product copy with AI and need it to read like a real person wrote it. After you specify or imply Chinese versus English, the skill picks an intensity—moderate for obvious tells only, heavy as the default for active rewrites, or full for a from-scratch human voice while preserving facts. It follows a short procedure: infer language, confirm level, then apply the matching reference checklist for that locale. Use it when a newsletter, blog post, landing section, or doc feels templated, when stakeholders flag “ChatGPT voice,” or before you publish growth content. It does not replace strategy, fact-checking, or SEO tooling; it is editorial de-roboting with explicit control over how aggressive the edit is.883installs3Image GenerationImage Generation routes your agent through the right image stack depending on runtime: on Codex, it steers toward the native image tool with a consistent editorial style—soft modern palette, flat illustration, clear hierarchy—then ensures deliverables live in your repo instead of a transient default output folder. When users name destinations or need project assets, the skill requires move/copy into the workspace with predictable naming. If built-in generation is not viable, it documents a Gemini API fallback driven by a Python script and environment key, keeping the same visual brief unless the user overrides style. Solo builders use it while drafting docs, landing hero art, or launch graphics without maintaining a separate design toolchain for every article.882installs4Content RewriteContent Rewrite is a marketing agent skill that takes a single source piece—blog post, announcement, or article—and produces separate versions tuned to LinkedIn, X (Twitter), Reddit, an English-language blog, or WeChat Official Account. It is built for solo builders and indie marketers who ship product updates but do not have a copy team to reformat the same story five different ways. Before writing, the skill requires you to lock perspective (personal first person vs objective third person) and select target platforms, because voice and credibility shift materially on professional networks versus Reddit. Each destination pulls from a dedicated reference guide covering tone, length, structure, and platform norms such as X’s character ceiling and Reddit’s anti-spam expectations. Use it when you already have core messaging and need channel-ready drafts for organic distribution rather than net-new research or SEO architecture. The output is adapted copy you can paste or lightly edit—not automated publishing or analytics.881installs5Chrome AutomationChrome Automation is a setup-oriented agent skill that documents how to install agent-browser and attach it to your everyday Chrome instance through the Chrome DevTools Protocol. Solo builders who ship browser-backed checks, form flows, or light scraping from Claude Code, Cursor, or Codex get a repeatable install path instead of guessing ports, profiles, and security toggles. The guide contrasts enabling remote debugging from the Chrome UI—recommended on Chrome 136 and newer—with launching Chrome from the command line when you control the profile. Connection modes include auto-discovery on 127.0.0.1:9222 and explicit CDP URLs when debugging runs elsewhere. It is not a full test suite; it is the plumbing layer that lets agent-browser drive tabs once debugging is on. Expect shell commands, npm global installs, and local browser permissions. After setup, agents can open URLs and interact with pages the same way you would manually, which supports validate prototypes, ship smoke tests, and operate on staging UIs without maintaining a separate Puppeteer-only project for every task.859installs6Raw Video ProcessingRaw Video Processing is a small Python + FFmpeg workflow skill for solo builders who record demos, podcasts-on-video, or Loom-style clips and need silence removed without manual timeline scrubbing. You point it at an input MP4 (or similar), tune silence detection (threshold, minimum gap length, optional padding), and get a shorter file with awkward pauses stripped out. It fits the content leg of the solo journey when you are polishing assets for YouTube, newsletters with video, or course modules—not when you are building product UI. Expect intermediate comfort with a terminal, FFmpeg installed, and reasonable defaults experimentation on your mic room noise. It is a script-style capability rather than an in-editor agent ritual: run locally, review the cut, then upload or drop into your usual NLE.833installs7Video To GifVideo to GIF is an agent skill for solo builders who have a short screen recording or clip and need a shareable animated asset without manually tuning FFmpeg. Instead of one-shot conversion, it encodes multiple quality and size tradeoffs—frame rate, width, and color depth—so you can compare outputs visually and pick what fits your channel or page weight budget. The workflow assumes FFmpeg and uv are on the machine; gifsicle is optional for additional lossy passes. Invoke it when you are unsure which parameters balance smooth motion against file size, or when you want a documented set of options before committing to one GIF for a launch post, tutorial, or product demo embed.833installs8Mermaid To ImageMermaid to Image is an agent skill that implements a small publishing workflow: find fenced Mermaid diagrams in Markdown (or similar text), render them to PNG through the public mermaid.ink API, write images into the project’s conventional assets folder, and swap code fences for standard image references. Solo builders who document system design in repo Markdown but deploy to GitHub Pages, Jekyll, or Dev.to—places that do not execute Mermaid in preview—use it to avoid maintaining duplicate diagram sources. The workflow begins by scoping targets (one file, a docs tree, or a specific block ordinal), inventorying blocks, then picking or creating an output directory consistent with the repo. It foregrounds user confirmation when multiple asset roots exist, which prevents broken relative paths in generated posts. Complexity stays intermediate because encoding, URL length limits, and path hygiene matter for large diagrams. The skill is multi-phase: primary shelf Build docs, but the same step appears when shipping changelogs and when launching blog mirrors of architecture content. Pattern is generator plus integration (HTTP to mermaid.ink), with network permission implied.826installs9Jupyter Notebook WritingJupyter Notebook Writing, as packaged here, is the Milvus Integration Tutorial Format Specification: a prescriptive style guide for bootcamp notebooks that teach vector-database integrations. Indie builders and technical writers contributing to milvus-io bootcamp—or mirroring that pattern for their own integration catalog—get one checklist for badges, headings, dependency blocks, and narrative sections so every tutorial reads and renders the same on GitHub and Colab. The skill enforces visual entry points (Open in Colab, View on GitHub), a clear title and introduction tying the provider to Milvus, a preparation section with upgraded pip installs and Colab-specific runtime reminders, then the hands-on integration body and closing guidance. Filename rules keep provider names recognizable in snake_case. It is phase-specific documentation work: you invoke it when drafting or refactoring a tutorial file, not when brainstorming product ideas. Agents benefit because the rules are concrete HTML snippets and section order, reducing editorial back-and-forth. Deliverables are spec-compliant markdown or notebook sources ready for conversion to ipynb and publication in the bootcamp tree.823installs10Screenshot CompressionScreenshot Compression is a Python CLI utility for solo builders who ship landing pages, blog posts, and app-store listings with heavy image folders. It walks paths you pass—individual files, a directory, or a recursive tree—and rewrites PNG and JPEG assets in place using the standard pngquant and jpegoptim binaries, so you do not need a separate export pipeline from Figma or Xcode simulators. Default quality targets balance visual fidelity with meaningful byte savings, and optional flags let you tighten or loosen compression per campaign. The script checks dependencies up front and exits with actionable install guidance instead of opaque subprocess errors, which keeps agent-driven runs predictable on fresh machines. Use it right before distribution or ASO uploads when you already have final screenshots but file sizes are bloated, especially when you are batch-updating dozens of locale or device frames.809installs11Md To Feishumd-to-feishu is an agent skill that turns a local Markdown file into a Feishu (Lark) document using the feishu-docx CLI, including uploading embedded images. Solo builders and small teams who standardize internal docs on Feishu use it when specs, runbooks, or marketing drafts already live in the repo and need a one-command publish without manual copy-paste. The agent derives or accepts a title, probes for uvx or an installed feishu-docx binary, and runs create with the file path—surfacing install steps when neither tool is present. It fits the Build docs phase for handoffs to collaborators in China or Feishu-first orgs, and pairs with Markdown authoring skills upstream.802installs12Github TrafficGitHub Traffic is a Python CLI tool that collects repository traffic metrics via the GitHub API and visualizes trends over customizable time periods. Solo builders use it to monitor repository engagement—tracking views, clones, and traffic sources to understand user interest. It matters for data-driven decisions about documentation, marketing, and feature prioritization based on real usage patterns.799installs13Mermaid To GifMermaid-to-GIF is a Python CLI skill for indie builders who already diagram flows in Mermaid and need motion assets for growth channels without a design team. It automates a pipeline that renders Mermaid via CDN, applies one of several animation styles in HTML, captures frames with Playwright, and encodes GIFs with FFmpeg. You can point it at a standalone .mmd file or scrape fenced mermaid blocks from markdown, which fits docs-first workflows common in solo SaaS and open-source projects. Default tuning (10 fps, 4s duration, progressive style) is sensible for social previews, while flags let you match brand background and resolution. The skill sits firmly in the content layer of the journey: you are not changing product architecture—you are making architecture visible. Expect to install FFmpeg and Python Playwright locally; the skill is operational glue rather than an in-editor generator, which keeps it portable across Claude Code, Cursor, and terminal-driven agents.797installs