技能 fetching-library-docs

📦

fetching-library-docs

Name: fetching-library-docs
Author: Asmayaseen

内容修订 r1 高风险 🌐 网络访问⚙️ 外部命令📁 文件系统访问

获取库 API 文档

开发者经常需要最新的 API 示例，但不想把完整文档加载到模型上下文中。此技能会获取 Context7 库文档，并返回聚焦的示例、签名和注释。

支持: Claude Codex Code(CC)

⚠️ 38 差

安装前请停下并请求确认。

请审查安装计划，并在更改文件前获得用户明确同意。

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "fetching-library-docs" from https://skillstore.io/skills/asmayaseen-fetching-library-docs.md and its manifest at https://skillstore.io/api/skills/asmayaseen-fetching-library-docs/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/asmayaseen-fetching-library-docs.md 已签名 manifest GET /api/skills/asmayaseen-fetching-library-docs/manifest 已签名 lockfile GET /api/skills/asmayaseen-fetching-library-docs/lockfile

测试它

正在使用“fetching-library-docs”。给我展示 React useState 示例。

预期结果:

一个简洁的响应，包含精选的 useState 示例、匹配的 API 签名，以及来自当前 Context7 文档的关键注释。

正在使用“fetching-library-docs”。 Next.js 路由是如何工作的？

预期结果:

一个简短的概念性概述，包含几个路由示例，并说明相关路由概念之间的联系。

正在使用“fetching-library-docs”。查找 Prisma 查询语法和重要警告。

预期结果:

对查询方法、代表性示例，以及关于安全或推荐用法的提取注释进行聚焦总结。

安全审计

高风险

v11 • 2026/7/9 打开版本化报告

The skill has a legitimate purpose: fetching and filtering library documentation from Context7. Most static findings are false positives from markdown examples, awk output, and expected shell pipeline use, but the bundled generic MCP client exposes direct shell execution through --stdio, and the helper scripts use npx package execution plus predictable /tmp PID files.

已扫描文件

1,364

分析行数

复核项

已忽略误报

已确认安全问题 (3)

高

scripts/mcp-client.py:204-207 scripts/mcp-client.py:421-423

User-Controlled Stdio Command Execution

The generic MCP client accepts a --stdio command and starts it with subprocess.Popen using shell=True. Direct use of scripts/mcp-client.py can execute arbitrary shell commands, even though the main skill workflow uses a fixed Context7 command.

The command string comes from the --stdio argument and is passed to Popen with shell=True. This is a direct local command execution path when the helper is used outside the fixed wrapper scripts.

中

scripts/start-server.sh:15 scripts/fetch-raw.sh:35-38

Automatic Package Execution with npx

The skill starts Context7 by running npx -y @upstash/context7-mcp. This downloads or executes a package during skill use, which creates supply-chain and network execution risk if the package source is compromised.

The scripts explicitly use npx -y to execute the Context7 MCP package. This appears intentional, but it is still a real external package execution risk.

中

scripts/start-server.sh:6-16 scripts/stop-server.sh:6-16

Predictable PID File Can Target Local Processes

The start and stop scripts store a process id in /tmp/context7-mcp-${PORT}.pid and later pass that value to kill. A stale or tampered PID file could cause the stop script to terminate an unrelated local process.

The PID file path is predictable and located in /tmp, and stop-server.sh trusts its contents. The impact is local process termination rather than data theft.

能力复核项 (21)

这些是真实的本地能力，对此技能可能是预期行为，因此需要复核，但不会按已确认恶意行为计分。

高

scripts/mcp-client.py:204

Python subprocess.Popen

self._process = subprocess.Popen(

subprocess.Popen starts the user-supplied --stdio command with shell=True. Direct use of the helper can execute arbitrary local shell commands.

中

scripts/start-server.sh:9

Shell command substitution

if [ -f "$PID_FILE" ] && kill -0 $(cat "$PID_FILE") 2>/dev/null; then

The script reads a PID from a predictable /tmp file and uses it in process checks or status output. This is local operational risk if the PID file is stale or tampered.

中

scripts/start-server.sh:10

Shell command substitution

echo "Context7 MCP already running on port $PORT (PID: $(cat $PID_FILE))"

The script reads a PID from a predictable /tmp file and uses it in process checks or status output. This is local operational risk if the PID file is stale or tampered.

中

scripts/start-server.sh:20

Shell command substitution

if kill -0 $(cat "$PID_FILE") 2>/dev/null; then

The script reads a PID from a predictable /tmp file and uses it in process checks or status output. This is local operational risk if the PID file is stale or tampered.

中

scripts/start-server.sh:21

Shell command substitution

echo "Context7 MCP started on port $PORT (PID: $(cat $PID_FILE))"

The script reads a PID from a predictable /tmp file and uses it in process checks or status output. This is local operational risk if the PID file is stale or tampered.

中

scripts/start-server.sh:6

Temp directory access

PID_FILE="/tmp/context7-mcp-${PORT}.pid"

The script writes a predictable PID file under /tmp. This can be tampered with by local users and affects later process-control decisions.

中

scripts/stop-server.sh:9

Shell command substitution

PID=$(cat "$PID_FILE")

The script trusts a PID read from a predictable /tmp file before terminating a process. A tampered PID file could target an unrelated local process.

中

scripts/stop-server.sh:12

Standard device file access

kill "$PID" 2>/dev/null

The stop script uses a predictable /tmp PID file or process-name matching to terminate local processes. This creates local process-control risk.

中

scripts/stop-server.sh:16

Standard device file access

kill -9 "$PID" 2>/dev/null || true

The stop script uses a predictable /tmp PID file or process-name matching to terminate local processes. This creates local process-control risk.

中

scripts/stop-server.sh:25

Standard device file access

pkill -f "context7-mcp.*--port.*${PORT}" 2>/dev/null && echo "Context7 MCP stopped" || echo "Context

The stop script uses a predictable /tmp PID file or process-name matching to terminate local processes. This creates local process-control risk.

中

scripts/stop-server.sh:6

Temp directory access

PID_FILE="/tmp/context7-mcp-${PORT}.pid"

The stop script uses a predictable /tmp PID file or process-name matching to terminate local processes. This creates local process-control risk.

低

scripts/mcp-client.py:146

HTTP client library

def request(self, method: str, params: Optional[dict] = None) -> dict: