Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问许可证需确认审计提醒

alchemyalchemy 效率

Agent Skill

alchemy 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。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。

总安装

906

周安装

37

GitHub Stars

公开资料未说明

下载量

290
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:alchemy(alchemy 效率)
来源仓库:https://github.com/georgejeffers/alchemy-skills
仓库路径:skills/alchemy
安装命令:
npx skills add https://github.com/georgejeffers/alchemy-skills --skill alchemy
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/georgejeffers/alchemy-skills --skill alchemy

简介

alchemy 用于查找、检索和筛选相关信息。

  • 适合根据关键词或任务场景快速定位候选结果。
  • 可通过 npx 命令从指定仓库安装使用。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。
  • 安装前应确认权限和维护状态,避免误触网络或文件操作。
  • alchemy 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Alchemy

Alchemy is a TypeScript-native Infrastructure-as-Code framework. This skill provides comprehensive knowledge of Alchemy's APIs, patterns, and conventions.

Load references/alchemy-concepts.md for full details on any topic below.

Quick Start

New Project

alchemy create my-app --template vite

Add to Existing Project

alchemy init

Core File: alchemy.run.ts

import alchemy from "alchemy";
import { Worker, KVNamespace } from "alchemy/cloudflare";

const app = await alchemy("my-app");

const kv = await KVNamespace("cache", { title: "my-cache" });

const worker = await Worker("api", {
  entrypoint: "./src/worker.ts",
  bindings: {
    CACHE: kv,
    API_KEY: alchemy.secret(process.env.API_KEY),
    STAGE: app.stage,
  },
});

await app.finalize();

Commands

alchemy deploy              # deploy (default stage = $USER)
alchemy deploy --stage prod # deploy to prod
alchemy dev                 # local dev with hot reload
alchemy destroy             # tear down all resources

Task Routing

Based on what the user asks, determine which knowledge to load:

User RequestWhat to Do
Set up Alchemy in a projectUse alchemy create or alchemy init, create alchemy.run.ts
Add a Workerimport {Worker} from "alchemy/cloudflare", configure with entrypoint + bindings
Add a database/KV/R2/queueImport from alchemy/cloudflare, create resource, bind to Worker
Use secretsalchemy.secret(process.env.X) for input, Secret.unwrap() in custom resources
Set up dev modealchemy dev, configure framework adapter/plugin
Use a framework (Vite/Astro/etc.)Load references/alchemy-concepts.md §11 for framework adapters
Deploy to productionalchemy deploy --stage prod, see CI guide patterns
Use Neon/Stripe/AWS/other providerLoad references/alchemy-concepts.md §13 for provider list
Build a custom resourceLoad references/resource-patterns.md for implementation patterns
Build a new provider for AlchemyRun the full Provider Development Workflow below

Key Concepts

Resources

Resources are memoized async functions with create/update/delete lifecycle. Every resource takes an ID and props:

const db = await D1Database("my-db", { name: "my-db" });

Bindings

Connect resources to Workers with type-safe bindings:

const worker = await Worker("api", {
  entrypoint: "./src/worker.ts",
  bindings: { DB: db, KV: kv, SECRET: alchemy.secret("value") },
});

Access in worker code with type inference:

import type { worker } from "../alchemy.run";
export default {
  async fetch(req: Request, env: typeof worker.Env) {
    await env.DB.prepare("SELECT 1").run();
  },
};

Secrets

alchemy.secret(process.env.API_KEY)  // wrap a value
alchemy.secret.env.API_KEY           // shorthand

Stages

Isolated copies of your infrastructure. Default is $USER locally:

alchemy deploy              # deploys to $USER stage
alchemy deploy --stage prod # deploys to prod stage

Dev Mode

Local emulation with Miniflare, hot reload, framework integration:

alchemy dev

Framework Adapters

Each framework has a resource and a Vite plugin/adapter:

FrameworkResourcePlugin/Adapter
ViteVitealchemy/cloudflare/vite
AstroAstroalchemy/cloudflare/astro
React RouterReactRouteralchemy/cloudflare/react-router
SvelteKitSvelteKitalchemy/cloudflare/sveltekit
NuxtNuxtalchemy/cloudflare/nuxt
TanStack StartTanStackStartalchemy/cloudflare/tanstack-start
BunSPABunSPA(no plugin needed)
Next.jsNextjsalchemy/cloudflare/nextjs
Redwood (RWSDK)Redwoodalchemy/cloudflare/rwsdk

Providers

Alchemy has 20+ providers. The main ones:

ProviderImportKey Resources
Cloudflarealchemy/cloudflareWorker, D1Database, KVNamespace, R2Bucket, Queue, DurableObjectNamespace, Hyperdrive, Zone, DnsRecords
Neonalchemy/neonNeonProject, NeonBranch, NeonDatabase, NeonRole
Stripealchemy/stripeWebhookEndpoint, Price, Product, Coupon, and 10+ more
AWSalchemy/awsFunction, Table, Vpc, Subnet, SecurityGroup, Bucket, Role
AWS Controlalchemy/aws/controlAWS.{Service}.{Resource} — covers full AWS CloudFormation
GitHubalchemy/githubGitHubSecret
Vercelalchemy/vercelVercelProject, VercelDnsRecord, VercelDeployment

Conventions

  • Default import: import alchemy from "alchemy"
  • Named imports from subpaths: import {Worker} from "alchemy/cloudflare"
  • await app.finalize() is always the last line of alchemy.run.ts
  • Physical names include app name + stage for uniqueness
  • Stage defaults to $USER locally; use --stage prod for production
  • Sensitive values always wrapped with alchemy.secret()
  • State stored in .alchemy/ directory (add to .gitignore or commit for shared state)

Reference Files

Load these on demand for detailed information:

  • references/alchemy-concepts.md — Comprehensive reference covering all Alchemy concepts: apps, stages, scopes, phases, resources, secrets, state, bindings, CLI, dev mode, profiles, framework adapters, serialization, and all 20+ providers with their resources.
  • references/resource-patterns.md — Implementation patterns for building custom resources (17 sections). Load when the user wants to create custom resources or contribute to Alchemy.
  • references/test-patterns.md — Test conventions and patterns (15 sections). Load when writing tests for custom resources.
  • references/doc-patterns.md — Documentation templates (10 sections). Load when writing docs for Alchemy providers.
  • references/checklist.md — Completeness verification checklist. Load when verifying a provider implementation.

Provider Development Workflow

Use this workflow when the user is building a new provider or custom resource for the Alchemy framework itself (not just using Alchemy in their app).

Phase 1: Research

Study the provider's API, identify CRUD-capable entities, authentication method, rate limits.

Phase 2: Design

Define resource list, Props/Output interfaces, API client approach, dependency order.

Phase 3: API Client

Create alchemy/src/{provider}/api.ts — minimal fetch wrapper, env var auth with Secret.unwrap() fallback.

Phase 4: Resources

Implement each resource. Load references/resource-patterns.md.

  • Resource("{provider}::{Resource}",...) with create/update/delete lifecycle
  • Physical name: props.name?? this.output?.name?? this.scope.createPhysicalName(id)
  • Must use function declaration (not arrow), context via this
  • Export type guard (is{Resource}) for binding support

Phase 5: Tests

Write tests. Load references/test-patterns.md.

  • alchemy.test(import.meta, {prefix: BRANCH_PREFIX})
  • try/finally with destroy(scope) and API verification

Phase 6: Documentation

Write docs, guide, examples. Load references/doc-patterns.md.

  • Resource docs: alchemy-web/src/content/docs/providers/{provider}/{resource}.md
  • Guide: alchemy-web/src/content/docs/guides/{provider}.mdx
  • Example: examples/{provider}/alchemy.run.ts

Phase 7: Verify

Run checklist. Load references/checklist.md.

  • bun format, bun vitest alchemy/test/{provider}/, bun tsc -b

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

32.46%
按下载量换算94

Claude

30.77%
按下载量换算89

Cursor

19.76%
按下载量换算57

Gemini CLI

9.68%
按下载量换算28

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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