Token导航 LogoToken导航

Agent Skill · 开发

lucia-auth(露西娅·奥思)

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。

lucia-auth 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为操作浏览器;许可证信息还不明确,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 14,329。

总安装 717周安装 29GitHub Stars 公开资料未说明下载 225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:lucia-auth(露西娅·奥思)
来源仓库:https://github.com/ameistad/agent-skills
仓库路径:skills/lucia-auth
安装命令:
npx skills add https://github.com/ameistad/agent-skills --skill lucia-auth
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/ameistad/agent-skills --skill lucia-auth

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

lucia-auth 用于辅助安全审计、权限检查和认证流程分析,适合梳理敏感配置和鉴权逻辑。适用于需要检查凭据风险、依赖安全或生成安全复核清单的开发场景。通过 npx skills add 命令从指定 GitHub 仓库安装,需确认最小权限和操作边界。

使用时不能将输出直接当作最终结论,涉及密钥或生产系统时应先脱敏并验证操作合法性。lucia-auth 属于开发类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/lucia-auth

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/ameistad/agent-skills --skill lucia-auth;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

    它需要操作浏览器,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

lucia-auth 怎么安装?

在终端执行 npx skills add https://github.com/ameistad/agent-skills --skill lucia-auth。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

lucia-auth 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。

lucia-auth 需要什么权限?

权限级别是操作浏览器。该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

lucia-auth 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。

lucia-auth 用的人多吗?

公开数据是累计安装 717 次,近一周安装 29 次,下载 225 次。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。

lucia-auth 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

  • 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。
  • 需要处理开发类任务、又不想每次都手写提示词的人。

不太适合

  • 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。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
lucia-auth

露西娅·奥思 · 本页

717—操作浏览器未明确通过

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

Lucia Auth

Implements secure, production-oriented authentication using the patterns from Lucia Auth and The Copenhagen Book.

Overview

For browser-based web apps, treat auth as a system instead of a pair of login routes. If a user asks to "add auth" and does not explicitly ask for a toy example, ship a production baseline by default.

Production Baseline For Web Apps

When the repo does not already provide equivalents, the default baseline is:

  1. Email/password sign up, sign in, and sign out
  2. Server-side sessions in HttpOnly, Secure, SameSite cookies
  3. Email verification plus resend flow
  4. Password reset flow
  5. Rate limiting on every auth and email-sending endpoint
  6. CSRF protection or strict Origin validation on every state-changing route
  7. Session invalidation on password change, email verification/change, and permission changes
  8. Re-authentication ("sudo mode") for sensitive actions such as changing password/email, deleting account, disabling MFA, or viewing API keys
  9. Safe post-auth redirects with open-redirect protection
  10. Optional OAuth, TOTP, or WebAuthn layered on top as requested

Do not default to JWTs in localStorage for a normal web app. Prefer server-side sessions unless the architecture clearly requires stateless tokens.

Required Reference Reads

For any browser-based web auth task, always read these first:

  • references/copenhagen-book/password-authentication.md
  • references/copenhagen-book/sessions.md
  • references/copenhagen-book/csrf.md
  • references/copenhagen-book/email-verification.md
  • references/copenhagen-book/password-reset.md
  • references/copenhagen-book/open-redirect.md
  • references/lucia/sessions-basic.md
  • references/lucia/sessions-inactivity-timeout.md
  • references/lucia/rate-limit-token-bucket.md

Read these when relevant:

  • OAuth: references/copenhagen-book/oauth.md, then references/lucia/tutorial-github-oauth.md or references/lucia/tutorial-google-oauth.md
  • MFA: references/copenhagen-book/mfa.md
  • WebAuthn / passkeys: references/copenhagen-book/webauthn.md, then references/lucia/example-email-password-2fa-webauthn.md
  • Session tradeoffs: references/lucia/sessions-overview.md, references/lucia/sessions-stateless-tokens.md

Implementation Workflow

  1. Inspect the repo for existing auth, user, email, session, middleware, and database patterns.
  2. Decide the auth surface area. For a standard web app, default to the production baseline above.
  3. Create or update the auth schema first. At minimum: user, session, email_verification, and password_reset. Add oauth_account, user_totp, user_recovery_code, or WebAuthn tables only when needed.
  4. Implement secure primitives next. Password hashing, session issuance/validation, CSRF/origin checks, rate limiting, token generation, and redirect validation.
  5. Wire complete user flows. Registration, login, logout, verify email, resend verification, forgot password, reset password, change password, and revoke sessions.
  6. Add sensitive-action re-auth. Changing email/password, account deletion, disabling MFA, or security settings should require a fresh session or a new credential challenge.
  7. Verify behavior. Confirm cookie flags, rate limits, invalidation behavior, expired-token handling, and unhappy paths.

Security Rules

Passwords

  • Hash passwords with Argon2id. Scrypt is an acceptable fallback, bcrypt only for legacy constraints.
  • Minimum length is 8 characters. Do not set a low maximum; 64-256 is a reasonable bound.
  • Do not silently trim, normalize, or truncate passwords.
  • Allow copy/paste and password managers.
  • Check leaked passwords with haveibeenpwned when practical.
  • Recommend MFA for security-sensitive apps and require it when the app meaningfully needs it.

Emails

  • Normalize emails to lowercase before storing or comparing.
  • Keep validation simple to avoid ReDoS.
  • Do not silently remove +tag segments.
  • If email is used for identity recovery, implement verification before treating it as trusted.

Sessions

  • Create a brand new session on login. Never reuse a pre-auth or anonymous session after authentication.
  • Store a hash of the session secret, not the raw secret.
  • Prefer idle expiration plus an absolute lifetime.
  • Track session freshness for re-auth flows.
  • Invalidate all sessions when credentials or privileges change.
  • Never accept session tokens through URLs.

Cookies

  • Use HttpOnly, Secure, Path=/, and SameSite=Lax by default.
  • Use SameSite=Strict only when the UX tradeoff is acceptable.
  • Refresh long-lived cookies when the server refreshes the session inactivity window.

CSRF

  • Do not mutate state with GET.
  • Check Origin on all non-GET browser requests and reject when missing or untrusted.
  • For form-heavy apps, add CSRF tokens as a second layer.
  • Keep CORS strict so CSRF tokens cannot be harvested cross-origin.

OAuth / OIDC

  • Use Authorization Code + PKCE.
  • Verify state and the code verifier on callback.
  • Prefer client_secret_basic when the provider supports it.
  • Link accounts by provider subject first, and only auto-link by email when the provider explicitly marks the email as verified.
  • Sanitize any post-login returnTo or redirect_to value to same-origin paths only.

Verification And Reset Tokens

  • Use high-entropy random tokens for links.
  • Hash link tokens before storage with SHA-256.
  • Keep tokens single-use and expire them automatically.
  • Put token-bearing pages behind Referrer-Policy: strict-origin.
  • Email-sending endpoints must be strictly rate limited.

Template Map

Use these assets as starting points and adapt them to the framework and ORM already in the repo:

  • Password hashing and password policy: assets/password-utils.ts
  • Session issuance, validation, inactivity timeout, and fresh-session helpers: assets/session-manager.ts
  • Core SQL schema: assets/session-schema.sql
  • Email verification links and codes: assets/email-verification.ts
  • Password reset flow: assets/password-reset.ts
  • OAuth + PKCE flow helpers: assets/oauth-handler.ts
  • CSRF and origin validation helpers: assets/csrf-protection.ts
  • Token bucket rate limiting: assets/rate-limit.ts
  • TOTP MFA and recovery codes: assets/totp-mfa.ts

When To Escalate Beyond The Baseline

Add WebAuthn or passkeys when:

  • The user explicitly asks for passkeys or WebAuthn
  • The product is security-sensitive
  • The app wants phishing-resistant MFA

Add OAuth when:

  • The product already uses external identity providers
  • The user explicitly asks for social login
  • The app benefits from lower signup friction

Use stateless tokens only when:

  • The app is an API-first system that cannot rely on same-origin cookies, or
  • There is a clear multi-service architecture reason

Output Expectations

When using this skill, the final answer should make it clear:

  1. Which auth flows were implemented
  2. Which security controls were included by default
  3. Any remaining assumptions or provider-specific configuration the user must supply
  4. What was verified locally and what still needs integration testing

平台分布

Codex

34.94%
按下载量换算79

Claude

30.75%
按下载量换算69

Cursor

16.27%
按下载量换算37

Gemini CLI

8.78%
按下载量换算20

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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