菜单
This skill should be used when the user asks to "convert this page to markdown", "extract this webpage", "save this article", "grab content from URL", "get markdown from this link", "scrape this page", provides a URL to extract, or wants clean web content without ads and clutter.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Recall, search, continue, or analyze past conversations. Triggers on recall phrases ("what did we discuss", "continue where we left off", "we decided"), retrospective phrases ("do a retro", "post-mortem", "what went well", "lessons learned", "find antipatterns"), and implicit signals (past-tense references, possessives without context, assumptive questions like "do you remember").
This skill should be used when the user asks to "design a CLI", "help me design command-line flags", "what flags should my tool have", "create a CLI spec", "refactor my CLI interface", "design a CLI my agent can call", or wants to design command-line UX (args/flags/subcommands/help/output/errors/config) before implementation or audit an existing CLI surface for consistency and composability.
Use for any image creation or editing request — logo, sticker, product mockup, nano banana, t2i, i2i, multi-reference compositing via generate.py. Not for HTML/CSS mockups, diagrams, or coded UI.
This skill should be used when the user says "update CLAUDE.md", "refresh CLAUDE.md", "sync CLAUDE.md with the codebase", "reorganize CLAUDE.md", "optimize project instructions", or when CLAUDE.md is stale, verbose, or out of sync.
Persist learnings to memory or maintain existing memories. Triggers on "extract learnings", "save this for next time", "remember this pattern", "consolidate memories", "dream", "clean up memories".
This skill should be used when the user says "get PR comments", "show PR feedback", "what comments on my PR", "PR review comments", "show me the review", "what did reviewers say", or asks about feedback on a pull request. Not for creating PRs or responding to comments.
| name | convert-to-markdown |
| description | This skill should be used when the user asks to "convert this page to markdown", "extract this webpage", "save this article", "grab content from URL", "get markdown from this link", "scrape this page", provides a URL to extract, or wants clean web content without ads and clutter. |
| argument-hint | <URL> |
| allowed-tools | ["Bash(ezycopy:*)","Bash(curl:*)","AskUserQuestion"] |
Extract clean markdown from any URL using the ezycopy CLI.
Default mode uses fast HTTP fetch. Add --browser when the page relies on client-side
JavaScript to render its content or when authentication cookies are required — the default
fetcher only sees the raw HTML response, not the JS-rendered DOM.
Common --browser cases: Twitter/X, single-page applications, paywalled content.
-c — copy output to clipboard-o <path> — save to file or directory--browser — use headless Chrome for JS-rendered or authenticated pages--no-images — strip image links-t <duration> — timeout (default: 30s)Run ezycopy <URL> [flags] with the chosen mode.
In --browser mode: run as a foreground process and do not redirect stderr with 2>&1.
Chrome outputs diagnostic messages to stderr that should flow naturally rather than
polluting stdout capture.
If the output is empty or suspiciously short and --browser was not used, retry with
--browser — the site likely requires JS rendering.
If ezycopy is not found, ask the user before installing:
curl -sSL https://raw.githubusercontent.com/gupsammy/EzyCopy/main/install.sh | sh
Present the extracted markdown to the user. If the user requested a file save, use -o.
If they requested clipboard, use -c. When no explicit destination was given, display the
content directly.