Token导航 LogoToken导航

Agent Skill · 开发

spider-weave(蜘蛛网)

spider-weave 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。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。

spider-weave 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,按安装量在开发类的 21,746 个 Skill 里排第 11,612。

总安装 1,665周安装 68GitHub Stars 4下载 533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:spider-weave(蜘蛛网)
来源仓库:https://github.com/autumnsgrove/groveengine
仓库路径:skills/spider-weave
安装命令:
npx skills add https://github.com/autumnsgrove/groveengine --skill spider-weave
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/autumnsgrove/groveengine --skill spider-weave

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。

用于处理 GitHub 仓库、Issue 和 Pull Request 信息。适合围绕仓库状态、代码变更或协作事项进行整理。可结合来源仓库和原始 README 核验具体用法。安装前建议确认权限范围和维护状态。spider-weave 属于开发类 Skill,可作为该场景下的辅助能力补充。

注意是否会触发联网、命令执行或文件读写操作。

来源仓库内路径:skills/spider-weave

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/autumnsgrove/groveengine --skill spider-weave;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

spider-weave 怎么安装?

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

spider-weave 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。

spider-weave 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。

spider-weave 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。

spider-weave 用的人多吗?

公开数据是累计安装 1,665 次,近一周安装 68 次,下载 533 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。

spider-weave 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。
  • 需要处理开发类任务、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。

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
spider-weave

蜘蛛网 · 本页

1,6654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

Spider Weave

The spider doesn't rush. It spins one thread at a time, anchoring each carefully before moving to the next. The web grows organically—radial strands first, then the spiral, each connection tested for strength. When complete, the web catches what matters while letting the wind pass through. Authentication woven this way is strong, resilient, and beautiful in its structure.

When to Activate

  • User asks to "add auth" or "set up authentication"
  • User says "protect this route" or "add login"
  • User calls /spider-weave or mentions spider/auth
  • Integrating OAuth (Google, GitHub, etc.)
  • Setting up session management
  • Protecting API routes
  • Adding role-based access control (RBAC)
  • Implementing PKCE flow
  • Connecting to Heartwood (GroveAuth)

Pair with: raccoon-audit for security review, beaver-build for auth testing


The Weave

SPIN --> CONNECT --> SECURE --> TEST --> BIND
  |         |          |          |        |
Create    Link       Harden    Verify   Lock In
Threads  Strands     Knots      Web    Security

Phase 1: SPIN

*The spider spins the first thread, anchoring it carefully in the corner of the frame...*

Create the foundational auth structure. Choose a pattern, scaffold the files, define the schema, and wire environment variables before writing a line of logic.

  • Choose the auth pattern: Session, JWT, OAuth 2.0, PKCE, or API Keys — based on app type and user needs
  • Scaffold src/lib/auth/ with index.ts, types.ts, session.ts, middleware.ts, pkce.ts, client.ts
  • Define the users and sessions tables in the database schema
  • Set required environment variables: client IDs, secrets, authorize/token/userinfo URLs, redirect URI, session secret

Reference: Load references/oauth-pkce-flow.md for PKCE setup code, database schema, env var list, and the full file structure


Phase 2: CONNECT

*Thread connects to thread, the web taking shape across the frame...*

Link the auth system together: implement the OAuth login redirect, the callback handler, user upsert, and session creation. Then wire up the client-side auth store.

  • Implement the login route: generate PKCE verifier + challenge, set cookies, redirect to provider
  • Implement the callback route: verify state, exchange code for tokens, fetch userinfo, upsert user, create session, clean up PKCE cookies
  • Build session creation (createSession), validation (validateSession), and invalidation (invalidateSession)
  • Set up the client-side auth store with loadUser() for reactive auth state in Svelte

Reference: Load references/oauth-pkce-flow.md for the complete login and callback route implementations, and references/session-management.md for session functions and the auth store


Phase 3: SECURE

*The spider tests each knot, tightening what holds loose, cutting what doesn't belong...*

Harden the authentication system before trusting it with users. Add route protection, security headers, CSRF validation, and rate limiting.

  • Implement requireAuth() middleware that validates sessions in hooks.server.ts
  • Implement requireRole(allowedRoles) for RBAC — 403 on unauthorized access
  • Set security headers on every response: X-Frame-Options, X-Content-Type-Options, Referrer-Policy, Content-Security-Policy
  • Add CSRF origin validation for state-changing endpoints
  • Add rate limiting: 5 attempts per 15 minutes per IP on auth endpoints

Reference: Load references/route-protection.md for middleware code and RBAC patterns, and references/security-headers.md for headers, CSRF, rate limiting, and Cloudflare edge rules


Phase 4: TEST

*The spider plucks the strands, verifying each vibrates true at the right frequency...*

Test authentication thoroughly across the happy path, security edge cases, and failure modes. Automated tests catch regressions; security tests verify the knots hold.

  • Test the full OAuth flow: redirects with PKCE challenge, callback handling, session creation
  • Test route protection: unauthenticated redirect, authenticated access, role enforcement
  • Test security properties: CSRF state rejection, secure cookie attributes, session fixation prevention
  • Cover failure modes: expired sessions, invalid codes, denied OAuth permissions

Reference: Load references/oauth-pkce-flow.md for OAuth flow tests and references/route-protection.md for route protection test suite


Phase 5: BIND

*The web is complete, every strand bound tight, the whole stronger than the sum of its threads...*

Finalize and lock in the authentication. Polish the user experience, configure monitoring, and produce the completion report.

  • Verify the integration checklist: login, logout, protected routes, session expiry, error handling, rate limiting, CSRF, headers, cookies
  • Add login UI loading and error states with proper aria-busy and role="alert" attributes
  • Configure auth event logging (userId, provider, IP — never tokens or passwords)
  • Alert on suspicious activity (10+ failed attempts from same identifier)
  • Generate the completion report: provider, flow type, files created, security features confirmed, test counts

Reference: Load references/session-management.md for the full integration checklist, login UI snippet, monitoring patterns, and completion report template. Load references/heartwood-integration.md for Grove-specific logout flow and role assignment.


Reference Routing Table

PhaseReferenceLoad When
SPINreferences/oauth-pkce-flow.mdAlways — foundation code lives here
CONNECTreferences/oauth-pkce-flow.md + references/session-management.mdImplementing login/callback/session
SECUREreferences/route-protection.md + references/security-headers.mdAdding middleware, headers, rate limiting
TESTreferences/oauth-pkce-flow.md + references/route-protection.mdWriting auth and route tests
BINDreferences/session-management.md + references/heartwood-integration.mdFinalizing, logging, reporting
Grove/Heartwoodreferences/heartwood-integration.mdAny Grove ecosystem integration

Spider Rules

Patience

Weave one thread at a time. Don't rush to connect everything at once. Each strand must be secure before adding the next.

Precision

Small mistakes in auth have big consequences. Verify every redirect, check every token, validate every session.

Completeness

A web with holes catches nothing. Test the error paths, the edge cases, the failure modes. Security is only as strong as the weakest strand.

Communication

Use weaving metaphors:

  • "Spinning the threads..." (creating foundations)
  • "Connecting the strands..." (linking components)
  • "Testing the knots..." (security hardening)
  • "The web holds..." (verification complete)

Anti-Patterns

The spider does NOT:

  • Store passwords in plain text (ever)
  • Skip PKCE in OAuth flows
  • Trust user input without validation
  • Leave default secrets in configuration
  • Ignore session expiration
  • Log sensitive data (tokens, passwords)
  • Reflect redirect_uri or next parameters without validation (open redirect)
  • Assign admin roles via OAuth claims (manual DB assignment only)

Example Weave

User: "Add GitHub OAuth login"

Spider flow:

  1. SPIN — "Create OAuth app in GitHub, generate client credentials, set up PKCE utilities, create auth endpoints structure"
  2. CONNECT — "Implement /auth/github/login redirect, /auth/github/callback handler, user upsert logic, session creation"
  3. SECURE — "Add CSRF state validation, secure cookie settings, rate limiting on auth endpoints, role assignment for new users"
  4. TEST — "Test OAuth flow, callback handling, session creation, protected route access, error cases (denied permissions)"
  5. BIND — "Add login button to UI, error state handling, loading states, documentation, monitoring"

Quick Decision Guide

SituationApproach
Simple app, internal usersSession-based auth
Public app, social loginOAuth 2.0 + PKCE
API for mobile/SPAJWT with refresh tokens
Service-to-serviceAPI keys with IP allowlist
Grove ecosystemHeartwood integration (load references/heartwood-integration.md)

Integration with Other Skills

Before Weaving:

  • eagle-architect — For auth system design decisions
  • swan-design — For auth flow specifications

During Weaving:

  • elephant-build — For multi-file auth implementation
  • raccoon-audit — For security review

After Weaving:

  • beaver-build — For auth testing
  • turtle-harden — For defense-in-depth hardening beyond the web
  • deer-sense — For accessibility audit of login UI

*A well-woven web catches intruders while letting friends pass through.*

平台分布

Codex

38.17%
按下载量换算203

Claude

31.6%
按下载量换算168

Cursor

18.61%
按下载量换算99

Gemini CLI

9.37%
按下载量换算50

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