What it does
Query your traces, logs, metrics and GitHub Actions CI runs in Everr with read-only SQL.
- Connects 在云端运行
- Account or key Sign in to the service when you connect (OAuth)
- Category 开发者工具
可搭配使用
教 AI 做这类工作的 Skill,可与本项一起使用
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
graphify
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
为什么排在这里
此项在“连接 AI 的应用”标签中,该标签不使用星标,按以下顺序排列
| 1. 在 Claude 连接器目录中 | 否 | |
|---|---|---|
| 2. 团队实测过 | 否 | |
| 3. 适合泰国用户 | 否 | |
| 4. 不计星标的使用量 | 20/100 | 热度 0 · 新鲜度 12 · 仍在响应 8 · 用户喜欢 +0 |
| 5. 登记日期 | 2026-11-13 |
用户喜欢: 0 · 最多加 5 分(每人 1 分),不能用钱买。
质量: 基础检查通过 4/4 项(仅展示,不用于排序)
安装
选择你使用的 AI 应用,然后复制设置。
用一句话让 AI 帮你安装
请帮我安装 Everr。按照 https://mcp.popnx.com/i/everr.md 操作,先把命令给我看,等我同意后再运行。粘贴到能在你电脑上工作的 AI(如 Claude Code、Cursor)。它会读取我们的步骤,先给你看命令,等你同意。无需密钥,无需注册。 · 查看 AI 会读取的全文
通过 4/4 项检查 · 你的提问会经过开发者的服务器(app.everr.dev) · ✓ 命令已检查 2026年11月13日
第一步:你用哪个 AI?
选择你用的应用,步骤会显示在下方。
可连接 Claude · 约 2 分钟
不需要密钥,不用安装程序 · 在 claude.ai 或桌面应用设置一次,即可在聊天中使用 · 免费账户可添加 1 个 · Team/Enterprise 方案由所有者(Owner)添加
打开 Claude 的 Connectors 页面
CustomizeConnectors
点击“Add” → “Custom” → “Web”,填写这 2 项
把这里的内容复制到 Claude 中同名的输入框,然后点击 Continue
Name · 名称EverrRemote MCP server URL · 地址https://app.everr.dev/mcp在 Claude 检测到的登录设置页点击 Continue · Authentication 按作者说明选择(Sign in now、Sign in when needed 或 No sign in),然后点击 Add
在聊天中开启
+ConnectorsEverr
试一试
用 Everr 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude 说正在使用 Everr,或回答中出现 Everr 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
能否连接 ChatGPT 取决于方案
完整 MCP:网页版 ChatGPT Business、Enterprise 或 Edu · Pro:以读取/获取权限连接,修改数据的工具可能用不了 · Plus:官方文档尚未确认,如果在插件页看到添加自定义 MCP 的选项可以试试 · 仅限网页
在网页版 ChatGPT 打开添加 MCP 的页面
Plugins+Add custom MCP server
填写名称和地址
输入名称,在 Connection 下粘贴地址,勾选风险提示,然后点击 Create as a plugin
Name · 名称EverrConnection · 地址https://app.everr.dev/mcp在聊天中选择 Everr
开启新聊天,输入前先从插件列表中选择 Everr
试一试
用 Everr 告诉我你能帮我做什么,然后做一件给我看✓ 如果 ChatGPT 说正在使用 Everr,或回答中出现 Everr 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 OpenAI 帮助页面核实 · 你的界面可能因版本略有不同
可连接 Claude Code · 约 2 分钟
不需要密钥,不用安装程序
打开终端
Mac:按 Command+空格,输入 Terminal · Windows:打开 PowerShell
粘贴这条命令并按 Enter
claude mcp add --transport http --scope user everr https://app.everr.dev/mcp在 Claude Code 中输入 /mcp
应能看到 Everr 在列表中并显示已连接
试一试
用 Everr 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude Code 说正在使用 Everr,或回答中出现 Everr 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
可连接 Cursor · 约 2 分钟
不需要密钥,不用安装程序
在 Cursor 中以 Agent 模式打开 AI 聊天
试一试
用 Everr 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Cursor 说正在使用 Everr,或回答中出现 Everr 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Cursor 帮助页面核实 · 你的界面可能因版本略有不同
可连接 VS Code · 约 2 分钟
不需要密钥,不用安装程序
在 VS Code 中以 Agent 模式打开 AI 聊天
试一试
用 Everr 告诉我你能帮我做什么,然后做一件给我看✓ 如果 VS Code 说正在使用 Everr,或回答中出现 Everr 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 VS Code 帮助页面核实 · 你的界面可能因版本略有不同
其他支持 MCP 的应用(如 Windsurf、Zed、Gemini CLI)添加方式各不相同,请在应用列表中查看
通过网页连接的应用使用的地址
https://app.everr.dev/mcpJSON 设置在下方“开发者专用”中
开发者专用:终端命令和 JSON 设置
通过网页连接
Claude Code
claude mcp add --transport http --scope user everr https://app.everr.dev/mcpCursor (~/.cursor/mcp.json)
{
"mcpServers": {
"everr": {
"url": "https://app.everr.dev/mcp"
}
}
}VS Code (mcp.json)
{
"servers": {
"everr": {
"type": "http",
"url": "https://app.everr.dev/mcp"
}
}
}这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
一览
哪些应用可以使用它
依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。
通过互联网连接,适用于
ClaudeChatGPTGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatOpen WebUIRaycast
它能做什么
服务器要求先登录,才会列出它的工具。
基础检查i
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月19日
- ✓ 通过
作者与仓库一致
注册名称已通过域名 everr.dev 验证(注册中心会核实域名所有权)
- ✓ 通过
仍在维护
代码最后更新于昨天
- – 不适用
有许可证
不适用:由 everr.dev 的所有者自己运行的云端服务器,你不会运行仓库里的代码
- ✓ 通过
端点有响应
在 603 毫秒内响应(HTTP 401)
- ✓ 通过
使用 OAuth 登录
声明支持 OAuth 登录,你的 AI 应用可以带你去登录
- … 尚未检查
工具描述中没有隐藏指令
尚未检查;我们会在检查端点时读取工具列表
云端地址最近 4 次检查中有 4 次响应 · 最近一次 603 ms · 检查于 2026年11月19日
作者:申请重新检查或报告信息错误
给作者:在 README 中加徽章
如果你是作者,可把这段代码贴到 README。徽章会链接回本页。
[](https://mcp.popnx.com/en/servers/everr)讨论
提问、分享技巧或报告无法使用。按“有帮助”数量排序。
用户评价
还没有评价
与“开发者工具”中的其他服务器比较i
- 星标数多于此分类中其他 6116 个服务器的 91%
- 发布到注册中心的时间比此分类中其他 6116 个服务器的 88% 更新
- 基础检查结果等于或优于此分类中其他 6116 个服务器的 99%
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
常见问题
Everr 安全吗?
它在 4 项适用的基础检查中通过了 4 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
它需要哪些密钥或代码?
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在我的电脑上运行还是在云端?
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
它还在维护吗?
GitHub 仓库最后更新于昨天,最新版本(1.0.0)于6天前进入注册中心。
哪些应用可以使用它?
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,因此也能用于接受远程 MCP 的应用,例如网页版 ChatGPT(Business 及以上为完整 MCP;Pro 为只读)。
有什么可以替代它?
同一分类中有:GitDiagram, Appllama, Cloudflare MCP, Omnara, Repowise
功能相近的其他选择
同一分类、其他作者、命令已通过检查。官方的排在前面,再按星标排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
GitDiagram
Architecture diagrams, explanations and Mermaid source for public GitHub repositories.
Appllama
Study screens, flows and paywalls from top-earning iOS apps, then build from what wins.
Omnara
Deploy durable AI agents with Omnara, an open-source managed-agent platform.
Repowise
Codebase intelligence for AI coding agents: graph, git history, docs, decisions, code health.
| 本项 Everr | GitDiagram | Appllama | Cloudflare MCP | Omnara | Repowise | |
|---|---|---|---|---|---|---|
| 运行位置 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 云端或安装在电脑上都可以 |
| 需要密钥 | 否 | 否 | 否 | 否 | 否 | 否 |
| 可在 Claude 中按 + 添加 | 否 | 否 | 否 | 是 | 否 | 否 |
| 基础检查 | 4/4 | 4/4 | 4/4 | 4/4 | 4/4 | 5/5 |
| 许可证 | AGPL-3.0 | MIT | MIT | Apache-2.0 | Apache-2.0 | AGPL-3.0 |
| 价格 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 |
作者的 README
打开 README
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Self-hosted application
@everr/app is the complete self-hosted application and the base for @everr/app-cloud. It owns shared product features and the defaults needed to run one workspace. Cloud composes those features and supplies its own policy and cloud capabilities. The app must remain usable without importing or depending on cloud.
For installation, configuration, deployment, and account recovery, see the deployment guide. For migration ownership and database compatibility, see DATABASE-MIGRATIONS.md.
Ownership
Shared features own their behavior and server-side authorization. Application specific decisions enter through typed contracts in src/composition, exported from @everr/app/composition. Each application provides adapters in its own src/app/composition directory. Keep contracts type-only so importing a contract does not load an application's implementation.
The self-hosted app owns the single-workspace installation flow, its default identity and retention policies, and the shared feature implementations. Cloud owns billing, usage metering, tenant provisioning, commercial policy, and its Polar integration. Shared code should depend on the narrow contract for the behavior it needs, not on cloud modules or application implementation paths.
Server responsibilities are separated by workflow: runtime manages migrations and workers; entry handles installation, signup, and guarded organization entry; deployment configuration provides the ingest endpoint; identity owns authentication and membership lifecycle behavior, including shared endpoint guards. Account preflight stays alongside identity. After membership validation, getActiveOrganizationAccess calls getOrganizationEntryRedirect(session) and applies the returned destination while preserving returnTo. Cloud owns organization creation cookies and continuation behavior. Identity adapters may implement provider.beforeMembershipMutation for lock acquisition and beforeMembershipChange or afterMembershipChange for application policy; creation-only organization hooks remain separate. Shared endpoint guards remain in place.
Identity composition selects two cohesive parts together: policy contains signup admission, account-deletion preflight, and membership/deletion rules; provider contains Better Auth persistence, plugins, organization-creation hooks, session onboarding, and request-wide locking. Signup policy receives an invitation ID rather than provider endpoint context. The identity module translates policy denials into provider responses and retains ownership of shared guards. Authentication is exposed through the explicitly server-only @everr/app/identity/server entry; browser identity and role helpers use @everr/app/identity/client and @everr/app/identity/roles. Keep provider internals inside src/identity.
createApplicationAuth in @everr/app/identity/module constructs identity from the selected policy/provider adapter, a database, authentication settings, mail delivery, and invitation eligibility. Importing this constructor does not read environment settings or create database clients. identity/server is the production composition entry. Database and request-lock constructors likewise accept explicit configuration or pools; application composition owns their instances. Tests can construct independent identities with disposable databases and external-provider doubles without replacing global modules.
Call auth.deleteAccount({ headers, deleteOrganization }) for account deletion. Identity resolves an uncached session and checks the authentication provider's freshness requirement before any deletion. It discovers organization IDs, acquires application-owned locks through provider.lockAccountDeletion, then reads current ownership and application policy while holding those locks through provider writes. Callers do not manage identity locks or account-deletion preflight. Organization deletion and user deletion are separate provider commits; if the second operation fails, the error explicitly reports that the organization was already deleted and preserves the original cause.
The provider's user.deleteUser.beforeDelete uses the same ownership and application-policy guard before credential or session cleanup begins. Server callers use auth.deleteAccount; HTTP deletion and token callbacks enter through the request-scoped auth.handler. Before discovering memberships, deletion takes provider.lockUserLifecycle. Shared organization creation takes the same lock, rechecks that the creator exists, and retains it through creation and its hooks. auth.api.createOrganization supplies the request scope for server callers, including onboarding and checkout callbacks. Cloud uses a per-user lock; the installation reuses its workspace lock. Contention fails immediately, so creation and billing can acquire their other locks without waiting in a cycle.
Expected identity policy denials throw IdentityPolicyError, exported from @everr/app/identity/policy-error. Identity translates them to forbidden provider responses. Application adapters translate their own business denials into this type and propagate database or external-provider failures unchanged.
Cloud organization creation, checkout completion, and downgrade are owned by cloud's organizations/commands.server.ts. Request handlers validate input and pass the authenticated actor to that workflow. The workflow owns eligibility, lock ordering, billing coordination, and persistence. Infrastructure exports such as database executors and migrations remain available for transactional extensions.
Shared membership mutations belong to @everr/app/identity/memberships. restrictOrganizationToOwner checks the retained Owner, removes other Members, and cancels pending invitations within the caller's transaction. Cloud owns downgrade eligibility and billing coordination and composes this shared operation with its plan update. The operation returns the actual removed Membership IDs for publication after that transaction commits; the shared module imports no cloud billing types.
Extending the application
- Use ordinary component props and
childrenwhen adding content to a shared component. For example, account composition keeps shared profile and session behavior while supplying the application's account-lifecycle action:
account: (props) => (
),
Keep shared authorization checks around actions restricted to administrators or signed-in users.
- Compose a page when an application needs different data loading or rendering.
ApplicationPages.memberscan provide both. Itsloadfunction may prefetch data, and its component owns rendering:
members: {
async load({ queryClient, organizationId }) { await queryClient.fetchQuery(membersPolicy(organizationId)); }, component: MembersPage, },
Shared routes continue to own authentication and authorization guards.
- Register new application routes explicitly in
AppComposition.routes. A route declaration names its mount point, URL path, and source file:
import { resolve } from "node:path";
import type { AppComposition } from "@everr/app/build/composition";
export const composition: AppComposition = { directory: import.meta.dirname, routes: { signedIn: [ { type: "route", path: "settings", file: resolve(import.meta.dirname, "src/routes/settings.tsx"), }, ], }, };
Shared routes retain file-based discovery. After changing routes, regenerate the application's checked-in TanStack route tree by running its development server or build. Do not add a component or feature registry for contributions.
- Keep application policy in its adapter. Shared identity code owns membership endpoint quirks and normalizes lifecycle behavior before invoking application hooks, so an adapter cannot replace the shared membership and deletion guards.
Each application's tsconfig.json selects a complete set of named @composition/* modules. Vite and Vitest use the same selection. The root pnpm run check command checks package imports and composition contracts. It also keeps identity policy free of authentication-provider dependencies, provider internals private, browser composition free of server implementations, and cloud request handlers free of direct database coordination. Browser pages may consume explicit server-function interfaces. These rules preserve the existing named adapters; new seams should correspond to behavior that actually varies.
Navigation uses the selected application's generated TanStack route types. Use literal route patterns with params for dynamic destinations. Shared navigation lists accept only parameter-free ApplicationPath values; RailRow preserves TanStack's route, params, and search inference. Breadcrumb segments accept a link element so each route can declare a fully typed directly. Compile-only navigation checks run with both applications' typecheck commands and verify that invalid routes, params, search, and cross-application links fail.
When changing migration behavior or supported upgrade paths, update the migration guide. When changing installation, deployment, or recovery procedures, update the deployment guide.
Local development
Follow the worktree development guide, then run from the repository root:
pnpm install
pnpm dev
Or use the interactive process interface:
pnpm dev --tui
App and Cloud run together with separate databases, users, and host collectors. The launcher supplies local connection settings and credentials. On first startup in an unconfigured worktree, it copies each application's .env from the main worktree. The applications load those files for integration settings. See .env.example for available settings.
Shared ClickHouse databases, base users, table definitions, and SQL query limits live in packages/app/clickhouse. A file tagged self-hosted or cloud after its number (0012-cloud-apply-rls.sql) applies to that product only. Each application applies them on startup to its own ClickHouse container, using the same app and otel database names. See DATABASE-MIGRATIONS.md. Each worktree has its own database volumes and Portless URLs. Ctrl+C stops the stack while retaining database data. In TUI mode, confirm the quit dialog.
Telemetry
Create an ingest key in the application and send OTLP/HTTP to its printed Portless ingest URL. App and Cloud use separate collectors and key-verification endpoints. The self-hosted App collector buffers in memory; Cloud retains persistent queues and usage accounting. Self-hosted reads do not filter by tenant.
Production self-hosted uses the separate deployment stack. Configure exporters with the installation's public ingestion URL and an ingest key.
Shared identity publishes User, Organization, and Membership lifecycle facts through EventPublisher . identity/identity-adapter.ts captures persisted before/after identity under locks and publishes after its outermost source transaction commits. Better Auth 1.6.33 also runs queued database after-hooks following rollback, so those hooks are not the publication boundary. Providers build adapters for the supplied database executor to preserve their billing and Organization lifecycle checks inside nested transactions. Installation setup publishes User, Organization, and initial Membership creation after its own transaction commits. Onboarding publishes its Organization update after its standalone statement succeeds. Membership facts retain the persisted relations