Token导航 LogoToken导航TokenDH.com
开发敏感数据github未标认证来源可访问许可证需确认审计异常

jwt-validateJWT validate 安全

Agent Skill

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

总安装

1,835

周安装

78

GitHub Stars

2

下载量

643
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

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:jwt-validate(JWT validate 安全)
来源仓库:https://github.com/jsonwebtoken/jwt-skills
仓库路径:skills/jwt-validate
安装命令:
npx skills add https://github.com/jsonwebtoken/jwt-skills --skill jwt-validate
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/jsonwebtoken/jwt-skills --skill jwt-validate

简介

jwt-validate 用于辅助安全审计、权限检查和认证流程分析。

  • 适合梳理敏感配置、检查依赖风险或生成安全复核清单。
  • 使用时不能把工具输出直接当最终结论。
  • 安装命令:npx skills add https://github.com/jsonwebtoken/jwt-skills --skill jwt-validate
  • 涉及密钥或用户数据时应确认最小权限和操作边界。

SKILL.md

JWT Validate

Verify a JWT's signature and validate its claims. Confirms the token is authentic, unexpired, and structurally sound.

Validation Order

Check in this order. Stop and report at the first failure.

1. Structure

  • Exactly 3 dot-separated parts, each valid base64url
  • Header and payload parse as valid JSON
  • Header contains alg; alg is not none (unless explicitly expected)

2. Claims

  • exp must be in the future (report time until expiry or how long ago it expired)
  • nbf must be in the past or present
  • iat must be in the past; flag if > 30 days old
  • iss, aud, sub — if user provides expected values, they must match
  • Allow 60 seconds clock skew tolerance on all time checks

3. Signature

Requires the user to provide a secret, PEM public key, or JWKS URI. Always pass secrets and tokens via inline env vars to avoid shell history exposure.

Node.js (preferred):

First, ensure jose is available — install it globally if missing:

node --input-type=module -e "await import('jose')" 2>/dev/null || npm install -g jose

Then verify the token:

JWT_TOKEN='the.jwt.here' JWT_SECRET='user-provided-secret' node --input-type=module -e "import {jwtVerify} from 'jose'; try { const {payload}=await jwtVerify(process.env.JWT_TOKEN, new TextEncoder().encode(process.env.JWT_SECRET), {algorithms:['HS256'],clockTolerance:60}); console.log('VALID'); console.log(JSON.stringify(payload,null,2)); } catch(e) { console.log('INVALID:',e.message); }"

Python:

JWT_TOKEN='the.jwt.here' JWT_SECRET='user-provided-secret' python3 -c "
import jwt,json,os
try:
  d=jwt.decode(os.environ['JWT_TOKEN'],os.environ['JWT_SECRET'],algorithms=['HS256'],leeway=60)
  print('VALID'); print(json.dumps(d,indent=2))
except Exception as e: print(f'INVALID: {e}')
"

JWKS verification (Node.js):

JWT_TOKEN='the.jwt.here' JWKS_URI='https://example.auth0.com/.well-known/jwks.json' node --input-type=module -e "import {jwtVerify,createRemoteJWKSet} from 'jose'; try { const {payload}=await jwtVerify(process.env.JWT_TOKEN, createRemoteJWKSet(new URL(process.env.JWKS_URI)), {algorithms:['RS256']}); console.log('VALID'); console.log(JSON.stringify(payload,null,2)); } catch(e) { console.log('INVALID:',e.message); }"

If no secret/key is provided, perform structure + claims validation only, and clearly state the signature was NOT verified.

Output Format

## JWT Validation Report
Structure: PASS
Claims:    PASS — exp 2025-06-15T12:00:00Z (expires in 2h)
Signature: PASS — RS256, key kid "abc123"
Result:    VALID

On failure:

## JWT Validation Report
Structure: PASS
Claims:    FAIL — exp 2024-01-15T12:00:00Z (expired 6 months ago)
Signature: SKIPPED
Result:    INVALID — token expired

Security Rules

  • Never trust the token's alg header for verification. Always use the algorithm the user expects or that matches the provided key type. Trusting the header enables algorithm confusion attacks where an attacker switches RS256 to HS256 and signs with the public key as an HMAC secret.
  • Always specify algorithms as an explicit allowlist in verification calls. Never pass algorithms: [decoded.header.alg].
  • Never pass secrets/tokens as literal command-line arguments. Use environment variables. Args are visible in shell history and ps output.
  • alg: none — Flag as a critical security issue. The token is unsigned and cannot be trusted.
  • If no key is provided, validate structure and claims only. Clearly state: "Signature was NOT verified — token authenticity is unknown."

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

37.32%
按下载量换算240

Claude

28.35%
按下载量换算182

Cursor

19.04%
按下载量换算122

Gemini CLI

9.23%
按下载量换算59

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

未通过

权限和风险

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。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills