菜单
Xiaolai's Claude tools collection. Use when user types /xiaolai. Routes to: (1) claude-agent-sdk — Agent SDK reference for building autonomous AI agents, or (2) nlpm — Natural-Language Programming Manager for scanning, scoring, and fixing NL artifacts.
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
Build autonomous AI agents with Claude Agent SDK. TypeScript v0.2.96 | Python v0.1.56. Covers: query(), hooks, subagents, MCP, permissions, sandbox, structured outputs, and sessions. Use when: building AI agents, configuring MCP servers, setting up permissions/hooks, using structured outputs, troubleshooting SDK errors, or working with subagents.
Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence. Use when asked to "open in browser", "test the site", "take a screenshot", or "dogfood this". (gstack)
OpenAI Codex CLI wrapper — three modes. Code review: independent diff review via codex review with pass/fail gate. Challenge: adversarial mode that tries to break your code. Consult: ask codex anything with session continuity for follow-ups. The "200 IQ autistic developer" second opinion. Use when asked to "codex review", "codex challenge", "ask codex", "second opinion", or "consult codex". (gstack) Voice triggers (speech-to-text aliases): "code x", "code ex", "get another opinion".
记录和跟踪个人重大决策。当用户说"记决策"、"记个决策"、"决策跟踪"、"回填"、"decision",或描述一个刚做的/正在考虑的重大决策(创业/事业/财务/人生)并希望系统化记录与复盘时使用。自动四层拆解(原始信息/决策事实/假设推断/结果验证),生成 2 个文件,维护索引与行为模式库,并在记新决策时用历史教训主动拦截。
Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence. Use when asked to "open in browser", "test the site", "take a screenshot", or "dogfood this". (gstack)
Restore working context saved earlier by /context-save. Loads the most recent saved state (across all branches by default) so you can pick up where you left off — even across Conductor workspace handoffs. Use when asked to "resume", "restore context", "where was I", or "pick up where I left off". Pair with /context-save. Formerly /checkpoint resume — renamed because Claude Code treats /checkpoint as a native rewind alias in current environments. (gstack)
| name | xiaolai |
| description | Xiaolai's Claude tools collection. Use when user types /xiaolai. Routes to: (1) claude-agent-sdk — Agent SDK reference for building autonomous AI agents, or (2) nlpm — Natural-Language Programming Manager for scanning, scoring, and fixing NL artifacts. |
| user-invocable | true |
You have two tools available. Ask the user which one they want to use:
/xiaolai sdk or /xiaolai agent)Build autonomous AI agents with Claude Agent SDK (TypeScript v0.2.96 / Python v0.1.56).
Covers: query(), hooks, subagents, MCP, permissions, sandbox, structured outputs, sessions.
Use when: building AI agents, configuring MCP servers, setting up permissions/hooks, using structured outputs, troubleshooting SDK errors, or working with subagents.
/xiaolai nlpm)Natural-Language Programming Manager — scan, lint, and score NL artifacts with Claude-native quality scoring.
Commands:
/nlpm:ls — discover NL artifacts/nlpm:score — 100-point quality scoring/nlpm:check — cross-component consistency/nlpm:fix — auto-fix mechanical issues/nlpm:trend — track score history/nlpm:test — run NL-TDD specs/nlpm:init — configure project/nlpm:security-scan — scan plugin for security risks/xiaolai with no argument, show the menu above and ask which tool to use./xiaolai sdk or /xiaolai agent — read claude-agent-sdk/SKILL.md and follow its instructions./xiaolai nlpm — read nlpm/CLAUDE.md and follow its instructions. For specific nlpm commands, read the corresponding file under nlpm/commands/.All files are relative to this skill's directory (~/.claude/skills/xiaolai/):
claude-agent-sdk/SKILL.md
claude-agent-sdk/SKILL-typescript.mdclaude-agent-sdk/SKILL-python.mdclaude-agent-sdk/templates/claude-agent-sdk/rules/nlpm/CLAUDE.md
nlpm/commands/*.mdnlpm/agents/*.mdnlpm/skills/nlpm/Both repos are git clones. To update, run:
cd ~/.claude/skills/xiaolai/claude-agent-sdk && git pull
cd ~/.claude/skills/xiaolai/nlpm && git pull