Token导航 LogoToken导航

Agent Skill · 研究检索

insecure-temp-files-anti-pattern(不安全的临时文件反模式)

insecure-temp-files-anti-pattern 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

insecure-temp-files-anti-pattern 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 69,971。

总安装 214周安装 9GitHub Stars 4下载 75核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:insecure-temp-files-anti-pattern(不安全的临时文件反模式)
来源仓库:https://github.com/igbuend/grimbard
仓库路径:skills/insecure-temp-files-anti-pattern
安装命令:
npx skills add https://github.com/igbuend/grimbard --skill insecure-temp-files-anti-pattern
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/igbuend/grimbard --skill insecure-temp-files-anti-pattern

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

该技能揭示不安全的临时文件创建所暴露的三类攻击媒介。适用于 Codex、Claude、Cursor、Gemini CLI 中的代码安全审查场景。提供符号链接攻击、权限漏洞和清理缺失的具体反模式示例。

安装命令:npx skills add https://github.com/igbuend/grimbard --skill insecure-temp-files-anti-pattern。需确认本地环境是否允许执行外部脚本或访问系统目录。

来源仓库内路径:skills/insecure-temp-files-anti-pattern

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/igbuend/grimbard --skill insecure-temp-files-a…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

insecure-temp-files-anti-pattern 怎么安装?

在终端执行 npx skills add https://github.com/igbuend/grimbard --skill insecure-temp-files-anti-patter…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

insecure-temp-files-anti-pattern 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

insecure-temp-files-anti-pattern 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

insecure-temp-files-anti-pattern 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

insecure-temp-files-anti-pattern 用的人多吗?

公开数据是累计安装 214 次,近一周安装 9 次,下载 75 次,GitHub 4 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

insecure-temp-files-anti-pattern 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
insecure-temp-files-anti-pattern

不安全的临时文件反模式 · 本页

2144接触敏感数据未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Insecure Temp Files Anti-Pattern

Severity: Medium

Summary

Insecure temporary file creation exposes three attack vectors: predictable file names enabling symlink attacks, insecure permissions allowing unauthorized access, and missing cleanup leaving sensitive data on disk. Attackers exploit these to read sensitive data, inject malicious content, or cause denial of service. AI-generated code frequently suggests simplistic file handling vulnerable to these attacks.

The Anti-Pattern

Never create temporary files without securing their location, naming, permissions, and lifecycle management.

1. Predictable File Names

Using a predictable name for a temporary file creates a race condition. An attacker can guess the file name and create a symbolic link (symlink) at that location pointing to a sensitive system file. When the application writes to its "temporary" file, it is actually overwriting the linked file.

BAD Code Example
# VULNERABLE: Predictable temporary file name in a shared directory.
import os

def process_user_data(user_id, data):
    # The filename is easy for an attacker to guess.
    temp_path = f"/tmp/userdata_{user_id}.txt"

    # Attacker's action (done before this code runs):
    # ln -s /etc/passwd /tmp/userdata_123.txt

    # When the application writes to the temp file for user 123,
    # it is actually overwriting the system's password file.
    with open(temp_path, "w") as f:
        f.write(data)

    # ... processing logic ...
    os.remove(temp_path)
GOOD Code Example
# SECURE: Use a library function that creates a securely named temporary file.
import tempfile

def process_user_data(user_id, data):
    # `tempfile.mkstemp()` creates a temporary file with a random, unpredictable name
    # and returns a low-level file handle and the path.
    # It also ensures the file is created with secure permissions (0600 on Unix).
    fd, temp_path = tempfile.mkstemp(prefix="userdata_", suffix=".txt")
    try:
        with os.fdopen(fd, 'w') as f:
            f.write(data)
        # ... processing logic ...
    finally:
        # Always ensure the file is cleaned up.
        os.remove(temp_path)

2. Insecure Permissions and Missing Cleanup

Creating a temporary file with default permissions can make it world-readable, allowing other users on the system to access its contents. Failing to delete the temporary file after use means that sensitive data may be left behind on the disk.

BAD Code Example
# VULNERABLE: World-readable permissions and no cleanup.
import uuid

def generate_report(data):
    # The name is random, but the permissions are not secure.
    temp_path = f"/tmp/{uuid.uuid4()}.pdf"

    # `open` with mode 'w' often uses default permissions like 0644,
    # which means other users on the system can read the file.
    with open(temp_path, "w") as f:
        f.write(data) # Sensitive report data is written.

    return temp_path # The path is returned, but the file is never deleted.
GOOD Code Example
# SECURE: Guaranteed cleanup using a context manager.
import tempfile

def generate_report(data):
    # `NamedTemporaryFile` creates a file that is automatically deleted
    # when the context manager is exited.
    with tempfile.NamedTemporaryFile(mode='w', suffix='.pdf', delete=True) as temp_f:
        # The file has a secure name and permissions.
        temp_f.write(data)
        temp_f.flush()

        # You can use `temp_f.name` to get the path and pass it to other functions.
        result = send_file_to_storage(temp_f.name)

    # The temporary file is automatically and reliably deleted here,
    # even if an error occurs inside the `with` block.
    return result

Language-Specific Examples

JavaScript/Node.js:

// VULNERABLE: Predictable name and no cleanup
const fs = require('fs');
const path = require('path');

function processUpload(userId, data) {
  const tempPath = `/tmp/upload_${userId}.dat`; // Predictable!
  fs.writeFileSync(tempPath, data); // World-readable by default
  // ... processing ...
  // File never deleted!
  return tempPath;
}
// SECURE: Use tmp module with automatic cleanup
const tmp = require('tmp');

function processUpload(userId, data) {
  // Creates file with mode 0600 (owner read/write only)
  const tempFile = tmp.fileSync({ prefix: 'upload-', postfix: '.dat' });

  try {
    fs.writeFileSync(tempFile.name, data);
    // ... processing ...
    return processFile(tempFile.name);
  } finally {
    tempFile.removeCallback(); // Guaranteed cleanup
  }
}

Java:

// VULNERABLE: Predictable name in shared directory
public void processData(String userId, byte[] data) throws IOException {
    File tempFile = new File("/tmp/data_" + userId + ".tmp"); // Predictable!
    Files.write(tempFile.toPath(), data); // Default permissions may be insecure
    // ... processing ...
    // No cleanup - file persists!
}
// SECURE: Use Files.createTempFile with try-with-resources
import java.nio.file.*;
import java.nio.file.attribute.*;

public void processData(String userId, byte[] data) throws IOException {
    // Create with restricted permissions (owner only)
    Set<PosixFilePermission> perms = PosixFilePermissions.fromString("rw-------");
    FileAttribute<Set<PosixFilePermission>> attr =
        PosixFilePermissions.asFileAttribute(perms);

    Path tempFile = Files.createTempFile("data-", ".tmp", attr);

    try {
        Files.write(tempFile, data);
        // ... processing ...
    } finally {
        Files.deleteIfExists(tempFile); // Guaranteed cleanup
    }
}

Go:

// VULNERABLE: Predictable path and missing cleanup
func processData(userID string, data []byte) error {
    tempPath := fmt.Sprintf("/tmp/data_%s.tmp", userID) // Predictable!
    if err := os.WriteFile(tempPath, data, 0644); err != nil { // World-readable!
        return err
    }
    // ... processing ...
    // No cleanup!
    return nil
}
// SECURE: Use os.CreateTemp with defer cleanup
import "os"

func processData(userID string, data []byte) error {
    // Creates file with mode 0600 automatically
    tempFile, err := os.CreateTemp("", "data-*.tmp")
    if err != nil {
        return err
    }
    defer os.Remove(tempFile.Name()) // Guaranteed cleanup
    defer tempFile.Close()

    if _, err := tempFile.Write(data); err != nil {
        return err
    }

    // ... processing ...
    return nil
}

Detection

  • Search for insecure temp directories: Grep for hardcoded temp paths:

- rg 'open\s*\(\s*["\']/(tmp|var/tmp)/' - rg 'File\.createTempFile|mktemp|tmpfile' (check if used correctly)

  • Identify predictable file names: Find patterns based on user IDs or timestamps:

- rg 'f"/tmp/{user_id}' 'f"/tmp/{username}' - rg 'new File\("/tmp/" \+ userId'

  • Check file permissions: Audit permission settings:

- rg 'os\.chmod.*0o[67]' (world-readable/writable) - Review code for missing os.umask(0o077) or tempfile usage

  • Verify cleanup logic: Ensure files are always deleted:

- rg 'open\(' | rg -v 'with|try.*finally|NamedTemporaryFile' - Check for missing defer f.Close() (Go) or using (C#)

Prevention

  • Use a trusted library for creating temporary files, such as tempfile in Python or Files.createTempFile in Java. These libraries are designed to handle naming and permissions securely.
  • Never construct temporary file paths using predictable names.
  • Ensure temporary files are created with restrictive permissions (e.g., only readable and writable by the owner, 0600).
  • Always clean up temporary files. Use try...finally blocks or language features like context managers (with in Python) to guarantee deletion.
  • Consider using in-memory buffers (like io.BytesIO in Python) instead of temporary files if the data is small enough to fit in memory.

Related Security Patterns & Anti-Patterns

References

平台分布

Codex

38.86%
按下载量换算29

Claude

30.57%
按下载量换算23

Cursor

18.78%
按下载量换算14

Gemini CLI

9.19%
按下载量换算7

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills