joeseesun avatar

Qiaomu Opencli Explorer

This is a comprehensive guide for building OpenCLI adapters that turn websites into command-line interfaces. It walks you through discovering APIs via browser DevTools, choosing authentication strategies across five tiers (from public APIs to full UI automation), and writing TypeScript adapters using the cli() API. The workflow is methodical: open the site in a browser, interact with elements to trigger lazy-loaded endpoints, capture network requests, then write code against confirmed APIs. What stands out is the emphasis on active exploration rather than static analysis, with concrete examples from Bilibili, Reddit, and Xiaohongshu showing how different sites require different approaches (Wbi signatures, Pinia store actions, XHR interception). It includes helper utilities, common pitfalls, and a decision tree for picking the right authentication tier based on what actually works.

skills.sh installs
1K
GitHub stars
998
last updated
Aug 21, 2026

Install

$ npx -y skills add joeseesun/qiaomu-opencli-skills --skill qiaomu-opencli-explorer --agent claude-code

Installs into .claude/skills of the current project.

Files

CLI-EXPLORER — 适配器探索式开发完全指南

本文档教你(或 AI Agent)如何为 OpenCLI 添加一个新网站的命令。
从零到发布,覆盖 API 发现、方案选择、适配器编写、测试验证全流程。

[!TIP] 只想为一个具体页面快速生成一个命令? 看 opencli-oneshot skill(~150 行,4 步搞定)。 本文档适合从零探索一个新站点的完整流程。

[!TIP] 想要全自动生成? 直接运行 opencli generate <url> [--goal <goal>],内部会自动走完 explore → synthesize → cascade → verify 全流程。 返回 SkillOutput(success / blocked / needs-human-check)。如果自动生成失败或需要人工介入,再按本文档的手动流程继续。


AI Agent 开发者必读:用浏览器探索

[!CAUTION] 你(AI Agent)必须通过浏览器打开目标网站去探索!
不要只靠 opencli explore 命令或静态分析来发现 API。
你拥有浏览器工具,必须主动用它们浏览网页、观察网络请求、模拟用户交互。

为什么?

很多 API 是懒加载的(用户必须点击某个按钮/标签才会触发网络请求)。字幕、评论、关注列表等深层数据不会在页面首次加载时出现在 Network 面板中。如果你不主动去浏览和交互页面,你永远发现不了这些 API。

AI Agent 探索工作流(必须遵循)

Sponsors

Details

Repository
github.com/joeseesun/qiaomu-opencli-skills
First seen
Jun 3, 2026

More from joeseesun/qiaomu-opencli-skills

All 6

Recommended

1juliusbrussee avatarcavemanjuliusbrussee/cavemanUltra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels618.3Kskills.sh installs
2mattpocock avatargrill-memattpocock/skillsA relentless interview to sharpen a plan or design.1.5Mskills.sh installs
3shadcn avatarimproveshadcn/improveSurvey any codebase as a senior advisor and produce prioritized, self-contained implementation plans for OTHER models/agents to execute. Strictly read-only on s40Kskills.sh installs
4obra avatarsystematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes326.7Kskills.sh installs
5forrestchang avatarkarpathy-guidelinesforrestchang/andrej-karpathy-skillsBehavioral guidelines to reduce common LLM coding mistakes through explicit assumptions, simplicity, and verifiable success criteria.17.9Kskills.sh installs
6vercel-labs avatarfind-skillsvercel-labs/skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express int4.3Mskills.sh installs