security-patterns

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Security Patterns

安全模式

Comprehensive security patterns for building hardened applications. Each category has individual rule files in

rules/

loaded on-demand.

用于构建高安全性应用的全面安全模式。每个分类在

rules/

目录下都有独立的规则文件，可按需加载。

Quick Reference

快速参考

Category	Rules	Impact	When to Use
Authentication	3	CRITICAL	JWT tokens, OAuth 2.1/PKCE, RBAC/permissions
Defense-in-Depth	2	CRITICAL	Multi-layer security, zero-trust architecture
Input Validation	3	HIGH	Schema validation (Zod/Pydantic), output encoding, file uploads
OWASP Top 10	2	CRITICAL	Injection prevention, broken authentication fixes
LLM Safety	3	HIGH	Prompt injection defense, output guardrails, content filtering
PII Masking	2	HIGH	PII detection/redaction with Presidio, Langfuse, LLM Guard
Scanning	3	HIGH	Dependency audit, SAST (Semgrep/Bandit), secret detection
Advanced Guardrails	2	CRITICAL	NeMo/Guardrails AI validators, red-teaming, OWASP LLM

Total: 20 rules across 8 categories

分类	规则数量	影响级别	使用场景
身份认证	3	关键	JWT令牌、OAuth 2.1/PKCE、RBAC/权限管理
深度防御	2	关键	多层安全架构、零信任架构
输入验证	3	高	Schema验证（Zod/Pydantic）、输出编码、文件上传
OWASP Top 10	2	关键	注入攻击预防、身份认证漏洞修复
LLM安全	3	高	提示注入防御、输出管控、内容过滤
PII掩码	2	高	结合Presidio、Langfuse、LLM Guard的PII检测/脱敏
扫描检测	3	高	依赖审计、SAST（Semgrep/Bandit）、密钥检测
高级管控	2	关键	NeMo/Guardrails AI验证器、红队测试、OWASP LLM合规

总计：8个分类共20条规则

Quick Start

快速开始

python

undefined

python

undefined

Argon2id password hashing

Argon2id密码哈希

from argon2 import PasswordHasher ph = PasswordHasher() password_hash = ph.hash(password) ph.verify(password_hash, password)


```python

from argon2 import PasswordHasher ph = PasswordHasher() password_hash = ph.hash(password) ph.verify(password_hash, password)


```python

JWT access token (15-min expiry)

JWT访问令牌（15分钟过期）

import jwt from datetime import datetime, timedelta, timezone payload = { 'sub': user_id, 'type': 'access', 'exp': datetime.now(timezone.utc) + timedelta(minutes=15), } token = jwt.encode(payload, SECRET_KEY, algorithm='HS256')


```typescript
// Zod v4 schema validation
import { z } from 'zod';
const UserSchema = z.object({
  email: z.string().email(),
  name: z.string().min(2).max(100),
  role: z.enum(['user', 'admin']).default('user'),
});
const result = UserSchema.safeParse(req.body);

python

undefined


```typescript
// Zod v4 Schema验证
import { z } from 'zod';
const UserSchema = z.object({
  email: z.string().email(),
  name: z.string().min(2).max(100),
  role: z.enum(['user', 'admin']).default('user'),
});
const result = UserSchema.safeParse(req.body);

python

undefined

PII masking with Langfuse

结合Langfuse的PII掩码

import re from langfuse import Langfuse

def mask_pii(data, **kwargs): if isinstance(data, str): data = re.sub(r'\b[\w.-]+@[\w.-]+.\w+\b', '[REDACTED_EMAIL]', data) data = re.sub(r'\b\d{3}-\d{2}-\d{4}\b', '[REDACTED_SSN]', data) return data

langfuse = Langfuse(mask=mask_pii)

undefined

import re from langfuse import Langfuse

def mask_pii(data, **kwargs): if isinstance(data, str): data = re.sub(r'\b[\w.-]+@[\w.-]+.\w+\b', '[REDACTED_EMAIL]', data) data = re.sub(r'\b\d{3}-\d{2}-\d{4}\b', '[REDACTED_SSN]', data) return data

langfuse = Langfuse(mask=mask_pii)

undefined

Authentication

身份认证

Secure authentication with OAuth 2.1, Passkeys/WebAuthn, JWT tokens, and role-based access control.

Rule	Description
`auth-jwt.md`	JWT creation, verification, expiry, refresh token rotation
`auth-oauth.md`	OAuth 2.1 with PKCE, DPoP, Passkeys/WebAuthn
`auth-rbac.md`	Role-based access control, permission decorators, MFA

Key Decisions: Argon2id > bcrypt | Access tokens 15 min | PKCE required | Passkeys > TOTP > SMS

采用OAuth 2.1、Passkeys/WebAuthn、JWT令牌及基于角色的访问控制实现安全身份认证。

规则	描述
`auth-jwt.md`	JWT创建、验证、过期处理、刷新令牌轮转
`auth-oauth.md`	结合PKCE、DPoP、Passkeys/WebAuthn的OAuth 2.1
`auth-rbac.md`	基于角色的访问控制、权限装饰器、多因素认证（MFA）

核心决策： Argon2id > bcrypt | 访问令牌有效期15分钟 | 强制要求PKCE | Passkeys > TOTP > 短信验证

Defense-in-Depth

深度防御

Multi-layer security architecture with no single point of failure.

Rule	Description
`defense-layers.md`	8-layer security architecture (edge to observability)
`defense-zero-trust.md`	Immutable request context, tenant isolation, audit logging

Key Decisions: Immutable dataclass context | Query-level tenant filtering | No IDs in LLM prompts

无单点故障的多层安全架构。

规则	描述
`defense-layers.md`	8层安全架构（从边缘到可观测性）
`defense-zero-trust.md`	不可变请求上下文、租户隔离、审计日志

核心决策： 不可变数据类上下文 | 查询级租户过滤 | LLM提示中不包含ID

Input Validation

输入验证

Validate and sanitize all untrusted input using Zod v4 and Pydantic.

Rule	Description
`validation-input.md`	Schema validation with Zod v4 and Pydantic, type coercion
`validation-output.md`	HTML sanitization, output encoding, XSS prevention
`validation-schemas.md`	Discriminated unions, file upload validation, URL allowlists

Key Decisions: Allowlist over blocklist | Server-side always | Validate magic bytes not extensions

使用Zod v4和Pydantic验证并清理所有不可信输入。

规则	描述
`validation-input.md`	采用Zod v4和Pydantic的Schema验证、类型转换
`validation-output.md`	HTML清理、输出编码、XSS预防
`validation-schemas.md`	可区分联合类型、文件上传验证、URL白名单

核心决策： 优先使用白名单而非黑名单 | 始终在服务端验证 | 验证魔术字节而非文件扩展名

OWASP Top 10

Protection against the most critical web application security risks.

Rule	Description
`owasp-injection.md`	SQL/command injection, parameterized queries, SSRF prevention
`owasp-broken-auth.md`	JWT algorithm confusion, CSRF protection, timing attacks

Key Decisions: Parameterized queries only | Hardcode JWT algorithm | SameSite=Strict cookies

针对最关键的Web应用安全风险的防护措施。

规则	描述
`owasp-injection.md`	SQL/命令注入预防、参数化查询、SSRF防护
`owasp-broken-auth.md`	JWT算法混淆防护、CSRF防护、时序攻击防护

核心决策： 仅使用参数化查询 | 硬编码JWT算法 | SameSite=Strict Cookie

LLM Safety

LLM安全

Security patterns for LLM integrations including context separation and output validation.

Rule	Description
`llm-prompt-injection.md`	Context separation, prompt auditing, forbidden patterns
`llm-guardrails.md`	Output validation pipeline: schema, grounding, safety, size
`llm-content-filtering.md`	Pre-LLM filtering, post-LLM attribution, three-phase pattern

Key Decisions: IDs flow around LLM, never through | Attribution is deterministic | Audit every prompt

适用于LLM集成的安全模式，包括上下文分离和输出验证。

规则	描述
`llm-prompt-injection.md`	上下文分离、提示审计、禁用模式
`llm-guardrails.md`	输出验证流水线：Schema、事实依据、安全、大小
`llm-content-filtering.md`	LLM前过滤、LLM后归因、三阶段模式

核心决策： ID绕开LLM传递，而非通过LLM | 归因具有确定性 | 审计所有提示

PII Masking

PII掩码

PII detection and masking for LLM observability pipelines and logging.

Rule	Description
`pii-detection.md`	Microsoft Presidio, regex patterns, LLM Guard Anonymize
`pii-redaction.md`	Langfuse mask callback, structlog/loguru processors, Vault deanonymization

Key Decisions: Presidio for enterprise | Replace with type tokens | Use mask callback at init

针对LLM可观测性流水线和日志的PII检测与掩码处理。

规则	描述
`pii-detection.md`	Microsoft Presidio、正则表达式模式、LLM Guard匿名化
`pii-redaction.md`	Langfuse掩码回调、structlog/loguru处理器、Vault去匿名化

核心决策： 企业级场景使用Presidio | 用类型令牌替换敏感数据 | 初始化时使用掩码回调

Scanning

扫描检测

Automated security scanning for dependencies, code, and secrets.

Rule	Description
`scanning-dependency.md`	npm audit, pip-audit, Trivy container scanning, CI gating
`scanning-sast.md`	Semgrep and Bandit static analysis, custom rules, pre-commit
`scanning-secrets.md`	Gitleaks, TruffleHog, detect-secrets with baseline management

Key Decisions: Pre-commit hooks for shift-left | Block on critical/high | Gitleaks + detect-secrets baseline

针对依赖、代码和密钥的自动化安全扫描。

规则	描述
`scanning-dependency.md`	npm audit、pip-audit、Trivy容器扫描、CI门禁
`scanning-sast.md`	Semgrep和Bandit静态分析、自定义规则、提交前检查
`scanning-secrets.md`	Gitleaks、TruffleHog、detect-secrets及基线管理

核心决策： 左移至提交前钩子 | 阻断关键/高危漏洞 | Gitleaks + detect-secrets基线

Advanced Guardrails

高级管控

Production LLM safety with NeMo Guardrails, Guardrails AI validators, and DeepTeam red-teaming.

Rule	Description
`guardrails-nemo.md`	NeMo Guardrails, Colang 2.0 flows, Guardrails AI validators, layered validation
`guardrails-llm-validation.md`	DeepTeam red-teaming (40+ vulnerabilities), OWASP LLM Top 10 compliance

Key Decisions: NeMo for flows, Guardrails AI for validators | Toxicity 0.5 threshold | Red-team pre-release + quarterly

采用NeMo Guardrails、Guardrails AI验证器和DeepTeam红队测试的生产级LLM安全措施。

规则	描述
`guardrails-nemo.md`	NeMo Guardrails、Colang 2.0流程、Guardrails AI验证器、分层验证
`guardrails-llm-validation.md`	DeepTeam红队测试（40+漏洞）、OWASP LLM Top 10合规

核心决策： NeMo用于流程，Guardrails AI用于验证器 | 毒性阈值0.5 | 发布前及每季度进行红队测试

Anti-Patterns (FORBIDDEN)

反模式（禁止使用）

python

undefined

python

undefined

Authentication

身份认证

user.password = request.form['password'] # Plaintext password storage response_type=token # Implicit OAuth grant (deprecated) return "Email not found" # Information disclosure

user.password = request.form['password'] # 明文存储密码 response_type=token # 隐式OAuth授权（已弃用） return "Email not found" # 信息泄露

Input Validation

输入验证

"SELECT * FROM users WHERE name = '" + name + "'" # SQL injection if (file.type === 'image/png') {...} # Trusting Content-Type header

"SELECT * FROM users WHERE name = '" + name + "'" # SQL注入 if (file.type === 'image/png') {...} # 信任Content-Type头

LLM Safety

LLM安全

prompt = f"Analyze for user {user_id}" # ID in prompt artifact.user_id = llm_output["user_id"] # Trusting LLM-generated IDs

prompt = f"Analyze for user {user_id}" # 提示中包含ID artifact.user_id = llm_output["user_id"] # 信任LLM生成的ID

PII

logger.info(f"User email: {user.email}") # Raw PII in logs langfuse.trace(input=raw_prompt) # Unmasked observability data

undefined

logger.info(f"User email: {user.email}") # 日志中包含原始PII langfuse.trace(input=raw_prompt) # 未掩码的可观测性数据

undefined

Detailed Documentation

详细文档

Resource	Description
references/oauth-2.1-passkeys.md	OAuth 2.1, PKCE, DPoP, Passkeys/WebAuthn
references/request-context-pattern.md	Immutable request context for identity flow
references/tenant-isolation.md	Tenant-scoped repository, vector/full-text search
references/audit-logging.md	Sanitized structured logging, compliance
references/zod-v4-api.md	Zod v4 types, coercion, transforms, refinements
references/vulnerability-demos.md	OWASP vulnerable vs secure code examples
references/context-separation.md	LLM context separation architecture
references/output-guardrails.md	Output validation pipeline implementation
references/pre-llm-filtering.md	Tenant-scoped retrieval, content extraction
references/post-llm-attribution.md	Deterministic attribution pattern
references/prompt-audit.md	Prompt audit patterns, safe prompt builder
references/presidio-integration.md	Microsoft Presidio setup, custom recognizers
references/langfuse-mask-callback.md	Langfuse SDK mask implementation
references/llm-guard-sanitization.md	LLM Guard Anonymize/Deanonymize with Vault
references/logging-redaction.md	structlog/loguru pre-logging redaction

资源	描述
references/oauth-2.1-passkeys.md	OAuth 2.1、PKCE、DPoP、Passkeys/WebAuthn
references/request-context-pattern.md	用于身份流转的不可变请求上下文
references/tenant-isolation.md	租户范围的存储库、向量/全文搜索
references/audit-logging.md	已清理的结构化日志、合规性
references/zod-v4-api.md	Zod v4类型、转换、变换、细化
references/vulnerability-demos.md	OWASP漏洞与安全代码示例对比
references/context-separation.md	LLM上下文分离架构
references/output-guardrails.md	输出验证流水线实现
references/pre-llm-filtering.md	租户范围的检索、内容提取
references/post-llm-attribution.md	确定性归因模式
references/prompt-audit.md	提示审计模式、安全提示构建器
references/presidio-integration.md	Microsoft Presidio设置、自定义识别器
references/langfuse-mask-callback.md	Langfuse SDK掩码实现
references/llm-guard-sanitization.md	结合Vault的LLM Guard匿名化/去匿名化
references/logging-redaction.md	structlog/loguru日志前脱敏

Related Skills

Capability Details

能力详情

authentication

Keywords: password, hashing, JWT, token, OAuth, PKCE, passkey, WebAuthn, RBAC, session Solves:

Implement secure authentication with modern standards
JWT token management with proper expiry
OAuth 2.1 with PKCE flow
Passkeys/WebAuthn registration and login
Role-based access control

关键词: password, hashing, JWT, token, OAuth, PKCE, passkey, WebAuthn, RBAC, session 解决问题:

采用现代标准实现安全身份认证
带合理过期时间的JWT令牌管理
结合PKCE的OAuth 2.1流程
Passkeys/WebAuthn注册与登录
基于角色的访问控制

defense-in-depth

Keywords: defense in depth, security layers, multi-layer, request context, tenant isolation Solves:

How to secure AI applications end-to-end
Implement 8-layer security architecture
Create immutable request context
Ensure tenant isolation at query level

关键词: defense in depth, security layers, multi-layer, request context, tenant isolation 解决问题:

如何端到端保护AI应用
实现8层安全架构
创建不可变请求上下文
在查询级确保租户隔离

input-validation

Keywords: schema, validate, Zod, Pydantic, sanitize, HTML, XSS, file upload Solves:

Validate input against schemas (Zod v4, Pydantic)
Prevent injection attacks with allowlists
Sanitize HTML and prevent XSS
Validate file uploads by magic bytes

关键词: schema, validate, Zod, Pydantic, sanitize, HTML, XSS, file upload 解决问题:

基于Schema验证输入（Zod v4、Pydantic）
用白名单预防注入攻击
清理HTML并预防XSS
通过魔术字节验证文件上传

owasp-top-10

Keywords: OWASP, sql injection, broken access control, CSRF, XSS, SSRF Solves:

Fix OWASP Top 10 vulnerabilities
Prevent SQL and command injection
Implement CSRF protection
Fix broken authentication

关键词: OWASP, sql injection, broken access control, CSRF, XSS, SSRF 解决问题:

修复OWASP Top 10漏洞
预防SQL和命令注入
实现CSRF防护
修复身份认证漏洞

llm-safety

Keywords: prompt injection, context separation, guardrails, hallucination, LLM output Solves:

Prevent prompt injection attacks
Implement context separation (IDs around LLM)
Validate LLM output with guardrail pipeline
Deterministic post-LLM attribution

关键词: prompt injection, context separation, guardrails, hallucination, LLM output 解决问题:

预防提示注入攻击
实现上下文分离（ID绕开LLM）
用管控流水线验证LLM输出
确定性LLM后归因

pii-masking

Keywords: PII, masking, Presidio, Langfuse, redact, GDPR, privacy Solves:

Detect and mask PII in LLM pipelines
Integrate masking with Langfuse observability
Implement pre-logging redaction
GDPR-compliant data handling

关键词: PII, masking, Presidio, Langfuse, redact, GDPR, privacy 解决问题:

在LLM流水线中检测并掩码PII
将掩码与Langfuse可观测性集成
实现日志前脱敏
GDPR合规的数据处理

security-patterns

Original

Translation

Security Patterns

安全模式

Quick Reference

快速参考

Quick Start

快速开始

Argon2id password hashing

Argon2id密码哈希

JWT access token (15-min expiry)

JWT访问令牌（15分钟过期）

PII masking with Langfuse

结合Langfuse的PII掩码

Authentication

身份认证

Defense-in-Depth

深度防御

Input Validation

输入验证

OWASP Top 10

OWASP Top 10

LLM Safety

LLM安全

PII Masking

PII掩码

Scanning

扫描检测

Advanced Guardrails

高级管控

Anti-Patterns (FORBIDDEN)

反模式（禁止使用）

Authentication

身份认证

Input Validation

输入验证

LLM Safety

LLM安全

PII

PII

Detailed Documentation

详细文档

Related Skills

相关技能

Capability Details

能力详情

authentication

authentication

defense-in-depth

defense-in-depth

input-validation

input-validation

owasp-top-10

owasp-top-10

llm-safety

llm-safety

pii-masking

pii-masking