跳到主要内容

guidance 精选

Answer user questions about QwenPaw installation and configuration: first locate and read local documentation, then distill the answer; if local information is insufficient, fall back to the official website documentation.

自动化 Codex CLIMCP 客户端 Apache-2.0 3.6 万
在 GitHub 打开

安装方式

cp -r src/qwenpaw/agents/skills/guidance-en ~/.claude/skills/guidance-en

把该 skill 目录放进客户端 skills 目录。不同客户端的目录可能不同,Claude Code 默认是 ~/.claude/skills/

整仓一次性获取:git clone --depth 1 https://github.com/agentscope-ai/QwenPaw.git

依赖与权限

  • 无需 API Key
  • 需要联网访问外部服务
  • 依赖外部命令:python3 awk sed
  • 以上为按正文与 frontmatter 自动推断的信号,仅供参考;安装前建议先阅读原文。

SKILL.md 正文

来源 agentscope-ai/QwenPaw@02b6a6d · 许可 Apache-2.0 · 内容未修改

QwenPaw Installation and Configuration Q&A Guide

Use this skill when the user asks about QwenPaw installation, initialization, environment configuration, dependency requirements, or common configuration options.

Core principles:

  • Check local documentation first, then answer
  • Base answers on what has actually been read, do not speculate
  • Answer in the same language the user used to ask

Standard Flow

Step 1: Locate the Documentation Directory

Use built-in path resolution (works for all install methods)

DOCS_DIR=$(python3 -c "from qwenpaw.constant import DOCS_DIR; print(DOCS_DIR or '')" 2>/dev/null)

If the above returns a non-empty path and the directory exists, use it directly and skip to Step 2.

If it fails (e.g., older version without DOCS_DIR), fall back in the following order:

Check for documentation directory in memory

First, check whether there is a documentation directory in memory. If found, use it directly; otherwise, proceed to the next step.

# Get the documentation directory from memory
DOCS_DIR=$(find ~/.qwenpaw/memory/ -type d -name "docs")

If there is no documentation directory in memory, continue with the following logic.

Check the documentation directory in the project source code

Run the following script logic to obtain the variable $QWENPAW_ROOT:

# Get the absolute path of the binary
QWENPAW_PATH=$(which qwenpaw 2>/dev/null || whereis qwenpaw | awk '{print $2}')

# Logical deduction: if the path contains .qwenpaw/bin/qwenpaw, the root is three levels up
# Example: /path/to/QwenPaw/.qwenpaw/bin/qwenpaw -> /path/to/QwenPaw
if [[ "$QWENPAW_PATH" == *".qwenpaw/bin/qwenpaw" ]]; then
    QWENPAW_ROOT=$(echo "$QWENPAW_PATH" | sed 's/\/\.qwenpaw\/bin\/qwenpaw//')
else
    # Fallback: try to get the parent of the parent directory
    QWENPAW_ROOT=$(dirname $(dirname "$QWENPAW_PATH") 2>/dev/null || echo ".")
fi

echo "Detected QwenPaw Root: $QWENPAW_ROOT"

Verify and list the documentation directory:

Use the derived $QWENPAW_ROOT to locate the documentation:

# Construct the standard documentation path
DOCS_DIR="$QWENPAW_ROOT/website/public/docs/"

# Check if the path exists and list files
if [ -d "$DOCS_DIR" ]; then
    find "$DOCS_DIR" -type f -name "*.md" | head -n 100
else
    # If the derived path is incorrect, perform a global fuzzy search
    find "$QWENPAW_ROOT" -type d -name "docs" | grep "website/public/docs"
fi

If project documentation does not exist, search the working directory

If documentation is still not found, search for available documentation content under the qwenpaw installation path:

# Look for characteristic files such as faq.en.md or config.zh.md
FILE_PATH=$(find . -type f -name "faq.en.md" -o -name "config.zh.md" | head -n 1)
if [ -n "$FILE_PATH" ]; then
    # Use dirname to get the directory containing the file
    DOCS_DIR=$(dirname "$FILE_PATH")
fi

If a documentation directory is found, save it in memory in this format:

# Documentation Directory
$DOCS_DIR = <doc_path>

Step 2: Documentation Search and Matching

Documentation files follow the naming format <topic>.<lang>.md (e.g., config.zh.md, config.en.md, quickstart.zh.md).

Use the find command to list all matching documents in the target directory, and identify the target as <doc_path> based on filename keywords (e.g., install, env, setup).

# List all matching documents
find $DOCS_DIR -type f -name "*.md"

If no suitable document is found, read all documentation contents in the next step.

Step 3: Read the Documentation Content

After finding candidate documents, read and identify the paragraphs relevant to the question. You can use:

  • cat <doc_path>
  • file_reader skill (recommended for longer documents or paginated reading)

If the documentation is long, prioritize reading the sections most relevant to the question (installation steps, configuration options, example commands, notes, version requirements).

Step 4: Extract Information and Respond

Extract key information from the documentation and organize it into an actionable answer:

  • Give the direct conclusion first
  • Then provide steps / commands / configuration examples
  • Include necessary prerequisites and common pitfalls

Language requirement: the answer language must match the language of the user's question (answer in Chinese if asked in Chinese, answer in English if asked in English).

Step 5 (Optional): Official Website Lookup

If the previous steps cannot be completed (no local documentation, missing documentation, or insufficient information), use the official website as a fallback:

  • http://qwenpaw.agentscope.io/

Answer based on the content available from the official website, and clearly state in the answer that the conclusion comes from the official website documentation.

Output Quality Requirements

  • Do not fabricate non-existent configuration options or commands
  • When there are version differences, clearly note "please refer to the current documentation version"
  • For paths, commands, and configuration keys, provide copy-pasteable original snippets whenever possible
  • If information is still insufficient, clearly state the gaps and tell the user what additional information is needed (e.g., operating system, installation method, error logs)

质量与来源

所属仓库
agentscope-ai/QwenPaw
仓库星标
3.6 万
最近更新
今天
采集时间
今天
commit
02b6a6d088e3
内容指纹
sha256:f07e74a897819ea
文件路径
src/qwenpaw/agents/skills/guidance-en/SKILL.md

同仓库其它 Skill

查看全部 48 个
edu-math-tutorial
数学题分步讲解视频的领域知识。适用场景:(1) 用户给出一道数学题并要求做讲解/ 解题视频;(2) 用户说"讲解这道题"、"生成解题视频"、"make a math tutorial"; (3) 需要中文数学教学视频,涉及方程、公式或几何图形。本 skill 只提供领域 知识(题目拆解、讲解节奏、旁白文案规范、版式与 KaTeX 排版风格);画面、 配音与成片一律用 Creator 原生 Element 与工具从源头构建。
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
professional-media-prompts
为 Creator 编译制作级视觉资产图、电影分镜图和参考生视频 Prompt。适用于 角色身份板/设定图、关键帧或多宫格分镜、短片动作节拍、R2V/I2V/T2V 提示词, 尤其适合要求角色一致性、镜头语言、参考图职责、Seedance 时间段或声音设计的 任务。不用于纯素材剪辑、只查询模型参数,或只评价已经生成的成片。
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
visual-asset-design
为角色/场景/道具设计视觉资产、编写或修复 VisualVariant 与阵容图的生成 Prompt 时使用。涵盖参考图布局(艺术身份板/序列关键帧/环境研究板/道具研究 板)、构图与镜头语言、图片审核误判规避、实体 continuity 与以主图为参考 的跨 Variant 衍生设计。不用于视频 Prompt 编译(见 professional-media-prompts)、 纯素材剪辑或只评价已生成产物。
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
alicloud_cli
阿里云 CLI 中文文档镜像检索与命令辅助:先走章节索引,再下钻正文页面,给出命令前必须有本地文档证据。
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
terraform-cli-setup
Terraform CLI 安装与初始化技能。当用户本地未安装 Terraform 时自动完成安装,确保 terraform 命令可用并能执行 init/validate。不负责 Provider 凭证配置,凭证在实际使用时由 terraform-skill 引导。
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
terraform-skill
Use when working with Terraform or OpenTofu - creating modules, writing tests (native test framework, Terratest), setting up CI/CD pipelines, reviewing configurations, choosing between testing approaches, debugging state issues, implementing security scanning (trivy, checkov), or making infrastructure-as-code architecture decisions
精选 agentscope-ai/QwenPaw 3.6 万 原文
Codex CLIMCP 客户端
agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI
presentation-structure
Knowledge about the presentation slide format, weight system, navigation, and section structure
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI
presentation-styling
Knowledge about CSS classes, component patterns, and syntax highlighting in the presentation
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI
vibe-to-agentic-framework
The conceptual framework behind the presentation — what "Vibe Coding to Agentic Engineering" means, why the journey is structured the way it is, and how every slide fits the narrative arc
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI
time-skill
Display the current time in Pakistan Standard Time (PKT, UTC+5). Use when the user asks for the current time, Pakistan time, or PKT.
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI
weather-fetcher
Instructions for fetching current weather temperature data for Dubai, UAE from Open-Meteo API
精选 shanraisshan/claude-code-best-practice 6.7 万 原文
Claude CodeCodex CLIGemini CLI