Mcp Percona Pg

io.github.dockndevai由 GitHub 账号 dockndevai 发布v0.3.3更新于 Oct 26, 2026

Manage Percona PostgreSQL + PgBouncer on Kubernetes — safe-by-default access modes and guards.

已验证STDIO仅桌面数据库云端与基础设施

概览

AI 生成的概览

通过 Percona Operator 查看 Kubernetes 上的 Percona PostgreSQL 与 PgBouncer 集群,默认只读并带受控的写操作。

可以做什么
  • 列出 kube-config 中的上下文与 PerconaPGCluster 资源,支持按命名空间过滤并标记受保护集群
  • 读取集群摘要、完整状态(Patroni 成员、就绪情况、条件)、连接端点与声明的用户名
  • 读取 PgBouncer 设置以及 Patroni 动态配置中调优的 PostgreSQL 参数
  • 列出备份与恢复资源及其状态和目标集群
  • 读写模式下可扩缩容、暂停、调参、切换扩展和创建备份;管理模式另有恢复、升级、提升和删除
适用场景
适用于在 Kubernetes 上用 Percona Operator for PostgreSQL v2 运行 PostgreSQL 的团队,希望助手回答集群健康、连接池、参数调优、备份与恢复相关问题,或在不手工编辑自定义资源的情况下做受控变更。
运行要求
通过 npx 从 npm 包本地运行,需要 Node.js。需要运行 Percona Operator for PostgreSQL v2 的 Kubernetes 集群,以及服务器可读取的 kube-config,由 KUBECONFIG_PATH 指定。PERCONA_MODE 选择只读、读写或管理模式,PERCONA_NAMESPACE 设置默认命名空间。另有可选的允许列表、受保护集群、各项操作开关、试运行和审计日志变量。
安装前请注意
默认只读,但读写和管理模式可以扩缩容、暂停、调参、备份、恢复、升级、提升或删除集群与备份。除非确实需要变更,否则保持 PERCONA_MODE 为只读,并配合命名空间与集群允许列表、受保护集群以及恢复、升级、删除的单独开关。kube-config 授予集群访问权限,应通过 RBAC 限制范围。密码不会被返回,但连接端点和用户名会。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Mcp Percona Pg,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

mcp-percona-pg

[CI] [License: MIT] [npm]

A Model Context Protocol server for the Percona Operator for PostgreSQL. It lets an MCP-capable client (Claude Desktop, Claude Code, Cursor, …) operate PostgreSQL + PgBouncer clusters on Kubernetes — topology, connection pooling, tuning, backups/PITR, DR, extensions, and lifecycle — with behaviour controlled entirely by flags.

It drives the operator's custom resources (PerconaPGCluster, PerconaPGBackup, PerconaPGRestore, PerconaPGUpgrade) through your kube-config, so the model works the way you already do: "scale dev-pg to 3 replicas", "switch pooling to transaction mode", "restore prod-pg to 12:00 UTC".

Safe by default: it starts read-only, can be scoped to an allowlist of namespaces and clusters, protects critical clusters from mutation, gates restore / upgrade / delete behind separate opt-ins, and requires typed confirmation for high-impact actions. It never reads or returns database credentials.

Features

  • Discovery & status — list clusters, per-cluster summary and raw .status (Patroni members, PostgreSQL/PgBouncer readiness), connection endpoints, backups and restores.
  • Connection pooling — read and update PgBouncer pool_mode and the global pool tunables (default_pool_size, max_client_conn, …).
  • PostgreSQL tuning — read/merge parameters via spec.patroni.dynamicConfiguration (the only Patroni-safe path).
  • Lifecycle — scale PostgreSQL/PgBouncer, pause/resume, toggle built-in extensions, on-demand backups.
  • DR & recovery — restore / point-in-time recovery, promote a standby, major-version upgrades — each individually gated.

Security model

LayerFlagEffect
Access modePERCONA_MODEread-only → read-write → admin; over-privileged tools are never registered
Namespace/cluster allowlistsPERCONA_NAMESPACE_ALLOWLIST, PERCONA_CLUSTER_ALLOWLISTscope what the agent can touch
Protected clustersPERCONA_PROTECTED_CLUSTERSreadable, never mutated/restored/deleted
Restore / upgrade / deletePERCONA_ALLOW_RESTORE, PERCONA_ALLOW_UPGRADE, PERCONA_ALLOW_DELETEseparate opt-ins on top of admin mode
ConfirmationPERCONA_REQUIRE_CONFIRMATIONhigh-impact ops require echoing the cluster name
Dry-run / auditPERCONA_DRY_RUN, PERCONA_AUDIT_LOGvalidate-only; JSON audit line per guarded op
Interactive confirmation(automatic)destructive & high-impact ops prompt the human to approve via MCP elicitation before running; fall back to the PERCONA_ALLOW_* gates when the client can't elicit

Tools

Read (read-only+): list_contexts, list_clusters, get_cluster, get_cluster_status, get_connection_info, get_pgbouncer_config, get_pg_parameters, list_backups, list_restores

Write (read-write+): scale_cluster, set_pgbouncer_config, set_pg_parameters, pause_cluster, toggle_builtin_extension, create_backup

Admin (admin): restore_cluster (needs PERCONA_ALLOW_RESTORE), upgrade_cluster (needs PERCONA_ALLOW_UPGRADE), promote_standby, delete_backup / delete_cluster (need PERCONA_ALLOW_DELETE)

Quickstart — add to your agent

Published on npm as @dockndevai/mcp-percona-pg. No clone or build needed — your MCP client runs it on demand with npx. Start in read-only mode; see .env.example for every variable and docs/CLIENTS.md for the full per-client guide.

Claude Code (CLI)

bash
claude mcp add percona-pg -e PERCONA_MODE="read-only" -e PERCONA_NAMESPACE="postgres-operator" -- npx -y @dockndevai/mcp-percona-pg

Claude Desktop · Cursor · Windsurf — same block in claude_desktop_config.json, .cursor/mcp.json, or ~/.codeium/windsurf/mcp_config.json:

json
{  "mcpServers": {    "percona-pg": {      "command": "npx",      "args": ["-y", "@dockndevai/mcp-percona-pg"],      "env": {        "PERCONA_MODE": "read-only",        "PERCONA_NAMESPACE": "postgres-operator"      }    }  }}

OpenAI Codex CLI — in ~/.codex/config.toml:

toml
[mcp_servers.percona-pg]command = "npx"args = ["-y", "@dockndevai/mcp-percona-pg"]env = { PERCONA_MODE = "read-only", PERCONA_NAMESPACE = "postgres-operator" }

Example prompts

  • "List the PostgreSQL clusters and show me the status of dev-pg."
  • "What pool_mode is dev-pg using, and how big is the default pool?" → get_pgbouncer_config
  • "Set dev-pg PgBouncer to transaction pooling with default_pool_size 25." (needs read-write)
  • "Bump shared_buffers to 512MB on dev-pg." (needs read-write)
  • "Take a full backup of dev-pg to repo1." (needs read-write)
  • "Restore dev-pg to 2026-08-30 12:00:00+00." (needs admin + PERCONA_ALLOW_RESTORE + confirmation)

Prerequisites

  • A Kubernetes cluster running the Percona Operator for PostgreSQL v2 (pgv2.percona.com/v2).
  • A kube-config the server can read. For safety, use a ServiceAccount/RBAC scoped to the operator's namespaces and to the pgv2.percona.com resources you want the agent to see.

Run from source (development)

Prefer the published package above. To run from a clone:

bash
npm installnpm run buildnode dist/index.js   # with the environment variables set

Develop

bash
npm run devnpm test          # security policy + annotationsnpm run typecheck

Publishing

This server ships a server.json for the official MCP registry and an mcpName for npm ownership validation. See PUBLISHING.md.

License

MIT

来源:README.md,提交 98d0aa5

工具

9

list_contexts

List kube-config contexts

List the contexts (clusters) available in the loaded kube-config.

无需输入参数。

list_clusters

List PostgreSQL clusters

List PerconaPGCluster resources (PostgreSQL + PgBouncer). Omit `namespace` to list across all namespaces. Results are filtered by the namespace/cluster allowlists; protected clusters are flagged.

输入参数contextnamespace

get_cluster

Get cluster (summary)

Summary of a PerconaPGCluster: state, PostgreSQL size, PgBouncer, version, standby, host.

输入参数name必填contextnamespace

get_cluster_status

Get cluster status (raw)

The full `.status` of a PerconaPGCluster — Patroni members, PostgreSQL/PgBouncer readiness, host, and conditions.

输入参数name必填contextnamespace

get_connection_info

Get connection info

Connection endpoints for a cluster: the primary/replica Service hosts, port, and the declared users (names only — passwords live in Secrets and are never returned).

输入参数name必填contextnamespace

get_pgbouncer_config

Get PgBouncer config

The PgBouncer settings for a cluster: replicas, pool_mode, and the global pool tunables.

输入参数name必填contextnamespace

get_pg_parameters

Get PostgreSQL parameters

The tuned PostgreSQL parameters from `spec.patroni.dynamicConfiguration.postgresql.parameters` (the only correct place to set them — direct postgresql.conf edits are reverted by Patroni).

输入参数name必填contextnamespace

list_backups

List backups

List PerconaPGBackup resources in a namespace, with their state, repo, and completion time.

输入参数contextnamespace

list_restores

List restores

List PerconaPGRestore resources in a namespace, with their target cluster and state.

输入参数contextnamespace

版本历史

3
  1. v0.3.3最新Oct 26, 2026
  2. v0.3.0Sep 29, 2026
  3. v0.2.3Sep 16, 2026

评分

已满足适用检查项的 90%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分
  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。

  • 已列出工具必需15 分
  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分