Skip to main content
Glama
Zabbix MCP Server

Zabbix MCP Server

by mhajder

Zabbix MCP Server

Zabbix MCP Server 是一个基于 Python 的模型上下文协议(MCP)服务器,旨在提供对 Zabbix 监控数据和管理功能的高级、可编程访问。它提供了一个现代 API,用于查询、自动化和集成 Zabbix 资源,如主机、模板、触发器、监控项、问题、事件、用户、代理、维护周期等。该服务器支持读写操作,具有强大的安全功能,适合与 AI 助手、自动化工具、仪表板和自定义监控工作流集成。

功能

核心功能

  • 使用灵活过滤查询 Zabbix 主机、模板、监控项、触发器和主机组

  • 按严重级别过滤检索问题、事件和告警

  • 访问监控项的历史和趋势数据

  • 监控触发器状态和问题严重级别

  • 管理维护周期和计划停机时间

  • 检索用户宏和配置数据

  • 获取 SLA 和服务信息

管理操作

  • 创建、更新和删除主机、模板和主机组(如果启用)

  • 管理触发器、监控项和发现规则

  • 配置维护周期和用户宏

  • 在受监控主机上执行脚本

  • 确认事件并关闭问题

  • 创建和管理用户和代理

  • 支持对主机和模板进行批量操作

高级功能

  • 速率限制和 API 安全功能

  • 只读模式,限制所有写操作以实现安全监控

  • 可选工具搜索转换,适用于大型工具目录

  • 用于 HTTP 传输的 Bearer 令牌认证

  • 全面的日志记录和审计跟踪

  • SSL/TLS 支持和可配置超时

  • 多种传输选项(STDIO、SSE、HTTP)

  • 可选的 Sentry 集成,用于错误跟踪

Related MCP server: mcp-zabbix

安装

先决条件

  • Python 3.11 到 3.14

  • 访问 Zabbix 服务器。支持 6.0 至 7.4 版本——这是 zabbix-utils 客户端接受的范围,除非设置 ZABBIX_SKIP_VERSION_CHECK=true,否则它会拒绝连接该范围之外的版本。

  • 有效的 Zabbix API 令牌或具有适当权限的用户凭据

从 PyPI 快速安装

最简单的入门方式是从 PyPI 安装:

# Using UV (recommended)
uvx zabbix-mcp

# Or using pip
pip install zabbix-mcp

在运行服务器之前,请记得为你的 Zabbix 实例配置环境变量:

# Create environment configuration
export ZABBIX_URL=https://zabbix.example.com/api_jsonrpc.php
export ZABBIX_TOKEN=your-zabbix-api-token

从源代码安装

  1. 克隆仓库:

git clone https://github.com/mhajder/zabbix-mcp.git
cd zabbix-mcp
  1. 安装依赖:

# Using UV (recommended)
uv sync

# Or using pip
pip install -e .
  1. 配置环境变量:

cp .env.example .env
# Edit .env with your Zabbix URL and credentials
  1. 运行服务器:

# Using UV (recommended)
uv run zabbix-mcp

# Or using the installed command directly
zabbix-mcp

使用 Docker

Docker 镜像可在 GitHub Packages 上获取,便于部署。

# Normal STDIO image
docker pull ghcr.io/mhajder/zabbix-mcp:latest

# MCPO image for usage with Open WebUI
docker pull ghcr.io/mhajder/zabbix-mcpo:latest

开发设置

用于使用附加工具进行开发:

# Clone and install with development dependencies
git clone https://github.com/mhajder/zabbix-mcp.git
cd zabbix-mcp
uv sync --group dev

# Run tests
uv run pytest

# Run with coverage
uv run pytest --cov=src/

# Run linting and formatting
uv run ruff check .
uv run ruff format .

# Run type checking
uv run ty check .

# Setup prek hooks
uv run prek install

配置

环境变量

# Zabbix Connection Details
ZABBIX_URL=https://zabbix.example.com/api_jsonrpc.php

# Authentication - use EITHER token OR user/password
# API Token (preferred over user/password)
ZABBIX_TOKEN=your-api-token
# OR Username/Password (for older versions)
# ZABBIX_USER=Admin
# ZABBIX_PASSWORD=zabbix

# SSL Configuration
ZABBIX_VERIFY_SSL=true
ZABBIX_TIMEOUT=30
ZABBIX_SKIP_VERSION_CHECK=false

# Read-Only Mode
# Set READ_ONLY_MODE true to disable all write operations (create, update, delete)
READ_ONLY_MODE=false

# Disabled Tags
# Comma-separated list of tags to disable tools for (empty by default)
# Example: DISABLED_TAGS=host,user,maintenance
DISABLED_TAGS=

# Logging Configuration
LOG_LEVEL=INFO

# Rate Limiting
# Set RATE_LIMIT_ENABLED true to enable rate limiting
RATE_LIMIT_ENABLED=false
RATE_LIMIT_MAX_REQUESTS=60
RATE_LIMIT_WINDOW_MINUTES=1

# Tool Search Transform (Optional)
# Set TOOL_SEARCH_ENABLED true to replace full tool listing with search tools
TOOL_SEARCH_ENABLED=false
# Search strategy: bm25 (natural language) or regex (pattern matching)
TOOL_SEARCH_STRATEGY=bm25
# Maximum number of matching tools returned by search_tools
TOOL_SEARCH_MAX_RESULTS=5

# Sentry Error Tracking (Optional)
# Set SENTRY_DSN to enable error tracking and performance monitoring
# SENTRY_DSN=https://your-key@o12345.ingest.us.sentry.io/6789
# Optional Sentry configuration
# SENTRY_TRACES_SAMPLE_RATE=1.0
# SENTRY_SEND_DEFAULT_PII=true
# SENTRY_ENVIRONMENT=production
# SENTRY_RELEASE=1.2.3
# SENTRY_PROFILE_SESSION_SAMPLE_RATE=1.0
# SENTRY_PROFILE_LIFECYCLE=trace
# SENTRY_ENABLE_LOGS=true

# MCP Transport Configuration
# Transport type: 'stdio' (default), 'sse' (Server-Sent Events), or 'http' (HTTP Streamable)
MCP_TRANSPORT=stdio

# HTTP Transport Settings (used when MCP_TRANSPORT=sse or MCP_TRANSPORT=http)
# Host to bind the HTTP server (default: 127.0.0.1)
# MCP_HTTP_HOST=127.0.0.1
# Port to bind the HTTP server (default: 8000)
# MCP_HTTP_PORT=8000
# Optional bearer token for authentication (leave empty for no auth)
# MCP_HTTP_BEARER_TOKEN=

可用工具

API 信息

  • api_version:获取 Zabbix API 版本信息

主机管理

  • host_get:列出主机,可按组、模板、代理和搜索条件进行可选过滤

  • host_create:创建具有接口和模板链接的新主机

  • host_update:更新主机属性(名称、状态、描述)

  • host_delete:按 ID 删除主机

主机组管理

  • hostgroup_get:列出主机组,可进行可选过滤

  • hostgroup_create:创建新的主机组

  • hostgroup_update:更新现有主机组的属性(名称)

  • hostgroup_delete:删除主机组

模板管理

  • template_get:列出模板,可进行可选过滤

  • template_create:创建新模板

  • template_update:更新模板属性(名称、描述)

  • template_delete:删除模板

监控项管理

  • item_get:列出监控项,可按主机、组、模板进行可选过滤

  • item_create:在主机上创建新的监控项

  • item_update:更新监控项属性(名称、延迟、单位、描述、状态)

  • item_delete:删除监控项

触发器管理

  • trigger_get:列出触发器,可按严重级别和状态过滤

  • trigger_create:使用表达式创建新触发器

  • trigger_update:更新触发器属性(描述、表达式、优先级、状态、注释)

  • trigger_delete:删除触发器

问题与事件管理

  • problem_get:获取当前问题,可按严重级别和时间过滤

  • event_get:获取事件,可按时间范围过滤

  • event_acknowledge:确认事件,可附带可选消息

历史与趋势

  • history_get:获取监控项的历史数据

  • trend_get:获取监控项的趋势数据

用户管理

  • user_get:列出用户,可进行可选过滤

  • user_create:创建新用户

  • user_update:更新用户属性(姓名、姓氏、密码、类型)

  • user_delete:删除用户

代理管理

  • proxy_get:列出代理,可进行可选过滤

  • proxy_create:创建新代理

  • proxy_update:更新代理属性(名称、运行模式、描述)

  • proxy_delete:删除代理

维护管理

  • maintenance_get:列出维护周期

  • maintenance_create:创建新的维护周期

  • maintenance_update:更新维护周期属性(名称、时间、描述)

  • maintenance_delete:删除维护周期

动作与媒体

  • action_get:列出动作(触发器、自动注册等)

  • mediatype_get:列出媒体类型

图形与发现

  • graph_get:列出图形,可进行可选过滤

  • discoveryrule_get:列出 LLD 发现规则

  • drule_get:列出网络发现规则

  • itemprototype_get:从发现规则获取监控项原型

SLA 与服务

  • sla_get:列出 SLA

  • service_get:列出服务

脚本

  • script_get:列出脚本

  • script_execute:在主机上执行脚本

用户宏

  • usermacro_get:列出用户宏(主机和全局)

  • usermacro_create:创建主机宏

  • usermacro_delete:删除主机宏

配置管理

  • configuration_export:将 Zabbix 配置导出为 JSON 或 XML

  • configuration_import:从 JSON 或 XML 导入 Zabbix 配置

安全与防护功能

只读模式

服务器支持只读模式,该模式禁用所有写操作以实现安全监控:

READ_ONLY_MODE=true

基于标签的工具过滤

你可以通过设置禁用标签来禁用特定类别的工具:

DISABLED_TAGS=alert,bills

大型工具集的工具搜索

FastMCP 工具搜索可以减少具有许多工具的服务器的提示大小。

启用后,list_tools 返回两个合成工具:

  • search_tools:查找匹配的工具并返回其完整模式

  • call_tool:按名称执行任何发现的工具

使用以下方式启用:

TOOL_SEARCH_ENABLED=true
TOOL_SEARCH_STRATEGY=bm25      # bm25 or regex
TOOL_SEARCH_MAX_RESULTS=8      # optional, default is 5

bm25 支持自然语言查询,而 regex 使用正则表达式 pattern 输入进行确定性匹配。

工具搜索遵循现有的可见性控制(只读模式和禁用标签)。

速率限制

服务器支持速率限制以控制 API 使用并防止滥用。如果启用,则使用滑动窗口算法对每个客户端的请求进行限制。

通过在 .env 文件中设置以下环境变量来启用速率限制:

RATE_LIMIT_ENABLED=true
RATE_LIMIT_MAX_REQUESTS=100   # Maximum requests allowed per window
RATE_LIMIT_WINDOW_MINUTES=1   # Window size in minutes

如果 RATE_LIMIT_ENABLED 设置为 true,服务器将应用速率限制中间件。根据你的环境调整 RATE_LIMIT_MAX_REQUESTS 和 RATE_LIMIT_WINDOW_MINUTES。

Sentry 错误跟踪与监控(可选)

服务器可选支持 Sentry 用于错误跟踪、性能监控和调试。Sentry 集成完全是可选的,仅在配置时初始化。

安装

要启用 Sentry 监控,请安装可选依赖:

# Using UV (recommended)
uv sync --extra sentry

配置

通过在 .env 文件中设置 SENTRY_DSN 环境变量来启用 Sentry:

# Required: Sentry DSN for your project
SENTRY_DSN=https://your-key@o12345.ingest.us.sentry.io/6789

# Optional: Performance monitoring sample rate (0.0-1.0, default: 1.0)
SENTRY_TRACES_SAMPLE_RATE=1.0

# Optional: Include personally identifiable information (default: true)
SENTRY_SEND_DEFAULT_PII=true

# Optional: Environment name (e.g., "production", "staging")
SENTRY_ENVIRONMENT=production

# Optional: Release version (auto-detected from package if not set)
SENTRY_RELEASE=1.2.3

# Optional: Profiling - continuous profiling sample rate (0.0-1.0, default: 1.0)
SENTRY_PROFILE_SESSION_SAMPLE_RATE=1.0

# Optional: Profiling - lifecycle mode for profiling (default: "trace")
# Options: "all", "continuation", "trace"
SENTRY_PROFILE_LIFECYCLE=trace

# Optional: Enable log capture as breadcrumbs and events (default: true)
SENTRY_ENABLE_LOGS=true

功能

启用后,Sentry 自动捕获:

  • 异常与错误:所有未处理异常及其完整上下文

  • 性能指标:请求/响应时间和跟踪

  • MCP 集成:详细的 MCP 服务器活动和交互

  • 日志与面包屑:用于调试的应用程序日志和事件轨迹

  • 上下文数据:环境、客户端信息和请求参数

获取 Sentry DSN

  1. 在 sentry.io 创建一个免费账户

  2. 创建一个新的 Python 项目

  3. 从项目设置中复制你的 DSN

  4. 将其设置到你的 .env 文件中

禁用 Sentry

Sentry 完全是可选的。如果你不设置 SENTRY_DSN,服务器将正常运行,不进行任何 Sentry 集成,也不会收集任何监控数据。

SSL/TLS 配置

服务器支持 SSL 证书验证和自定义超时设置:

ZABBIX_VERIFY_SSL=true    # Enable SSL certificate verification
ZABBIX_TIMEOUT=30         # Connection timeout in seconds

传输配置

服务器支持 MCP 协议的多种传输机制:

STDIO 传输(默认)

默认传输使用标准输入/输出进行通信。这非常适合本地使用以及与通过 stdin/stdout 通信的工具集成:

MCP_TRANSPORT=stdio

HTTP SSE 传输(服务器发送事件)

对于基于网络的部署,你可以使用带有服务器发送事件的 HTTP。这允许通过 HTTP 实时流式访问 MCP 服务器:

MCP_TRANSPORT=sse
MCP_HTTP_HOST=127.0.0.1        # Localhost
MCP_HTTP_PORT=8000           # Port to listen on
MCP_HTTP_BEARER_TOKEN=your-secret-token  # Optional authentication token

当使用带有 Bearer 令牌的 SSE 传输时,客户端必须在其请求中包含令牌:

curl -H "Authorization: Bearer your-secret-token" http://localhost:8000/sse

HTTP 可流式传输

HTTP 可流式传输提供基于 HTTP 的通信,支持请求/响应流式传输。这非常适合需要 HTTP 端点的 Web 集成和工具:

MCP_TRANSPORT=http
MCP_HTTP_HOST=127.0.0.1        # Localhost
MCP_HTTP_PORT=8000           # Port to listen on
MCP_HTTP_BEARER_TOKEN=your-secret-token  # Optional authentication token

当使用带有 Bearer 令牌的可流式传输时:

curl -H "Authorization: Bearer your-secret-token" \
     -H "Accept: application/json, text/event-stream" \
     -H "Content-Type: application/json" \
     -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' \
     http://localhost:8000/mcp

注意:HTTP 传输需要正确的 JSON-RPC 格式,包含 jsonrpc 和 id 字段。服务器可能还需要为某些操作进行会话初始化。

有关 FastMCP 传输的更多信息,请参阅 FastMCP 文档。

贡献

  1. 分叉仓库

  2. 创建功能分支(git checkout -b feature/amazing-feature)

  3. 进行你的更改

  4. 运行测试并确保代码质量(uv run pytest && uv run ruff check .)

  5. 提交你的更改(git commit -m 'Add amazing feature')

  6. 推送到分支(git push origin feature/amazing-feature)

  7. 打开拉取请求

许可证

MIT 许可证 - 有关详细信息,请参阅 LICENSE 文件。

Available Tools

53 tools
action_getAction GetB
Read-onlyIdempotent

Get actions from Zabbix.

Actions define automated responses to problems/triggers. They specify what happens when problems occur - sending notifications, executing remote commands, etc.

Inputs: limit, offset, output, search, +7 more. Returns: object

api_versionApi VersionA
Read-onlyIdempotent

Get Zabbix API version information.

This tool retrieves the current version of the Zabbix API you are connecting to. This is useful for understanding API capabilities and ensuring compatibility with specific features that may be version-dependent.

Inputs: no named parameters. Returns: object

configuration_exportConfiguration ExportA
Read-onlyIdempotent

Export Zabbix configurations.

Exports monitored hosts, templates, and their complete configurations to JSON, XML, or YAML format. Useful for backup, migration, disaster recovery, or sharing configurations.

When you export templates or hosts, the export includes:

  • All associated items (metrics/data sources)

  • All triggers and their dependencies

  • Discovery rules and prototypes

  • Graphs and visualizations

  • Macros and variable definitions

  • Host groups and interfaces (for hosts)

  • Inventory data (for hosts)

  • And all other configuration elements

Inputs: hostids, format_type, prettyprint, templateids. Returns: object

configuration_importConfiguration ImportB
Destructive

Import configurations into Zabbix.

Imports hosts, templates, and other configurations from JSON, XML or YAML. Useful for migration, cloning, or restoring configurations.

Inputs: content (required), rules, format_type, delete_missing. Returns: object

discoveryrule_getDiscoveryrule GetB
Read-onlyIdempotent

Get discovery rules from Zabbix.

Discovery rules automatically detect items, triggers, and interfaces from network resources. They enable dynamic host and item management without manual configuration.

Inputs: limit, offset, output, search, +7 more. Returns: object

drule_getDrule GetB
Read-onlyIdempotent

Get network discovery rules from Zabbix.

Network discovery (drule) rules perform network scanning to discover hosts and services. They can scan for active devices, open ports, and available services in CIDR ranges.

Inputs: limit, offset, output, search, +5 more. Returns: object

event_acknowledgeEvent AcknowledgeA
DestructiveIdempotent

Acknowledge events in Zabbix.

Mark events (problems/alerts) as acknowledged to show that operations staff are aware of and working on the issue. Acknowledged events can also be closed if resolved.

Inputs: eventids (required), action, message. Returns: object

event_getEvent GetA
Read-onlyIdempotent

Get events from Zabbix.

Events represent state changes in the system - when triggers transition from normal to problem and back, or recovery events. Each event has a timestamp, trigger, and can be acknowledged to show operators have seen the alert.

Inputs: limit, offset, output, hostids, +13 more. Returns: object

graph_getGraph GetC
Read-onlyIdempotent

Get graphs from Zabbix.

Graphs visualize item data over time, displaying metric values in line/bar/pie charts. Graphs can be included in dashboards, reports, and custom views for data analysis.

Inputs: limit, offset, output, search, +10 more. Returns: object

history_getHistory GetA
Read-onlyIdempotent

Get history data from Zabbix.

Retrieves the raw metric values collected by items. History contains all individual collected data points with timestamps, allowing detailed analysis of system behavior over time.

Inputs: itemids (required), limit, history, sortfield, +4 more. Returns: object

host_createHost CreateA

Create a new host in Zabbix.

Adds a new monitored host to Zabbix. This is essential for starting to monitor a new server or device. You must specify at least a host name and groups. You can optionally configure interfaces (for agent/SNMP communication) and link templates.

Inputs: host, name, groups, params, +4 more. Returns: object

host_deleteHost DeleteA
Destructive

Delete hosts from Zabbix.

Permanently removes one or more hosts from Zabbix. This will delete all associated data including history and alerts. Use with caution as this is a destructive operation.

Inputs: hostids (required). Returns: object

host_getHost GetA
Read-onlyIdempotent

Get hosts from Zabbix with optional filtering.

Retrieves a list of monitored hosts from Zabbix. You can filter by host IDs, groups, templates, proxies, or use search criteria. This is useful for discovering which hosts are available in your monitoring system.

Inputs: limit, offset, output, search, +14 more. Returns: object

hostgroup_createHostgroup CreateB

Create a new host group in Zabbix.

Host groups serve as containers for organizing hosts. They're essential for applying permissions, templates, and maintenance windows to multiple hosts at once.

Inputs: name (required). Returns: object

hostgroup_deleteHostgroup DeleteA
Destructive

Delete host groups from Zabbix.

Permanently removes one or more host groups. Hosts in deleted groups will no longer be members of that group (though the hosts themselves remain unless explicitly deleted).

Inputs: groupids (required). Returns: object

hostgroup_getHostgroup GetB
Read-onlyIdempotent

Get host groups from Zabbix.

Retrieves host groups with optional filtering. Host groups are used to organize and manage hosts collectively, applying templates, permissions, and maintenance windows to multiple hosts at once.

Inputs: limit, offset, output, search, +7 more. Returns: object

hostgroup_updateHostgroup UpdateA

Update an existing host group in Zabbix.

Modifies properties of an existing host group. You can change the group's name. Only specify the fields you want to change.

Inputs: groupid (required), name. Returns: object

host_updateHost UpdateA
DestructiveIdempotent

Update an existing host in Zabbix.

Modifies properties of an existing host. You can change the technical name, visible name, status (enable/disable monitoring), description, linked templates, or the proxy that monitors it. Only specify the fields you want to change.

Inputs: hostid (required), host, name, status, +4 more. Returns: object

item_createItem CreateB

Create a new item in Zabbix.

Items define what data is collected from a host. Each item specifies the metric name, how to collect it (agent, SNMP, etc.), what interval to use, and what data type to store.

Inputs: key_ (required), name (required), type_ (required), hostid (required), +5 more. Returns: object

item_deleteItem DeleteA
Destructive

Delete items from Zabbix.

Permanently removes one or more items from monitoring. The item's historical data is typically removed as part of cleanup, though this depends on Zabbix configuration.

Inputs: itemids (required). Returns: object

item_getItem GetB
Read-onlyIdempotent

Get items (metrics) from Zabbix.

Items are the data sources in Zabbix - they define what metrics are collected and how (protocol, interval, etc.). Each item produces a stream of values over time.

Inputs: limit, offset, output, search, +14 more. Returns: object

itemprototype_getItemprototype GetB
Read-onlyIdempotent

Get item prototypes from Zabbix.

Item prototypes are template items created by discovery rules that generate actual items dynamically based on discovered entities.

Inputs: limit, offset, output, search, +7 more. Returns: object

item_updateItem UpdateA

Update an existing item in Zabbix.

Modifies properties of an existing monitoring item. You can change the name, collection interval, units, or status. Only specify the fields you want to change.

Inputs: itemid (required), name, delay, units, +2 more. Returns: object

maintenance_createMaintenance CreateA

Create a new maintenance period in Zabbix.

Schedules a maintenance window when monitoring alerts are suppressed. Useful for planned upgrades, patching, or system maintenance without triggering false alarms.

Inputs: name (required), active_till (required), active_since (required), hostids, +3 more. Returns: object

maintenance_deleteMaintenance DeleteA
Destructive

Delete maintenance periods from Zabbix.

Cancels maintenance windows immediately, resuming alert generation. If the maintenance period has already passed, historical event suppression is retained.

Inputs: maintenanceids (required). Returns: object

maintenance_getMaintenance GetB
Read-onlyIdempotent

Get maintenance periods from Zabbix.

Maintenance windows define periods when monitoring is paused for planned upgrades, maintenance, or testing. Alerts are suppressed during maintenance periods.

Inputs: limit, offset, output, hostids, +5 more. Returns: object

maintenance_updateMaintenance UpdateB
Destructive

Update an existing maintenance period in Zabbix.

Modifies properties of an existing maintenance window. You can change the name, start time, end time, time periods, or description. Only specify the fields you want to change.

Inputs: maintenanceid (required), name, active_till, description, +2 more. Returns: object

mediatype_getMediatype GetB
Read-onlyIdempotent

Get media types from Zabbix.

Media types define communication channels for sending notifications (email, SMS, webhooks, etc.). Actions use media types to deliver alerts to users and integrations.

Inputs: limit, offset, output, search, +5 more. Returns: object

problem_getProblem GetB
Read-onlyIdempotent

Get problems from Zabbix.

Problems are active trigger states that indicate issues with monitored infrastructure. Each problem is associated with a trigger and can be acknowledged by operators.

Inputs: limit, offset, output, recent, +15 more. Returns: object

proxy_createProxy CreateA

Create a new proxy in Zabbix.

Proxies allow distributed monitoring by collecting data from remote networks and reporting to the central Zabbix server. Useful for firewall-separated networks or high-latency connections.

Inputs: name (required), description, operating_mode. Returns: object

proxy_deleteProxy DeleteA
Destructive

Delete proxies from Zabbix.

Permanently removes proxy definitions. Hosts assigned to deleted proxies will need to be reassigned to other proxies or the server. Data from deleted proxies is typically retained.

Inputs: proxyids (required). Returns: object

proxy_getProxy GetC
Read-onlyIdempotent

Get proxies from Zabbix.

Proxies act as data collection points for Zabbix, allowing monitoring of remote networks without direct connectivity. Proxies collect data locally and report to the Zabbix server.

Inputs: limit, offset, output, search, +5 more. Returns: object

proxy_updateProxy UpdateA

Update an existing proxy in Zabbix.

Modifies properties of an existing proxy. You can change the name, operating mode, or description. Only specify the fields you want to change.

Inputs: proxyid (required), name, description, operating_mode. Returns: object

script_executeScript ExecuteB
Destructive

Execute a script on a host in Zabbix.

Runs a custom script on a specified host. Used for executing remediation tasks, configuration changes, or diagnostic commands remotely.

Inputs: hostid (required), scriptid (required). Returns: object

script_getScript GetB
Read-onlyIdempotent

Get scripts from Zabbix.

Scripts are custom automation routines that can be executed on monitored hosts or the server. They can be triggered manually or by actions to automate remediation or configuration tasks.

Inputs: limit, offset, output, search, +7 more. Returns: object

service_getService GetB
Read-onlyIdempotent

Get services from Zabbix.

Services represent business capabilities or applications (e.g., 'Web Application', 'Database'). Services can depend on other services, creating hierarchies for tracking dependencies.

Inputs: limit, offset, output, search, +6 more. Returns: object

sla_getSla GetC
Read-onlyIdempotent

Get SLAs from Zabbix.

Service Level Agreements (SLAs) define uptime and availability targets for services. They track compliance with service objectives and generate reports on availability.

Inputs: limit, offset, output, search, +6 more. Returns: object

template_createTemplate CreateB

Create a new template in Zabbix.

Templates define the monitoring configuration (items, triggers, graphs) that can be reused across multiple hosts. Creating custom templates enables standardized monitoring for specific applications or server types.

Inputs: host (required), groups (required), name, description. Returns: object

template_deleteTemplate DeleteA
Destructive

Delete templates from Zabbix.

Permanently removes one or more templates. Hosts that have the deleted templates applied will lose those template's items, triggers, and graphs. The hosts themselves remain unchanged.

Inputs: templateids (required). Returns: object

template_getTemplate GetB
Read-onlyIdempotent

Get templates from Zabbix.

Templates are reusable collections of items, triggers, and graphs that can be applied to hosts. They standardize monitoring across multiple servers with the same role.

Inputs: limit, offset, output, search, +12 more. Returns: object

template_updateTemplate UpdateA

Update an existing template in Zabbix.

Modifies properties of an existing template. You can change the name or description. Only specify the fields you want to change.

Inputs: templateid (required), name, description. Returns: object

trend_getTrend GetA
Read-onlyIdempotent

Get trend data from Zabbix.

Trends are aggregated (summarized) historical data providing min/max/average values at hour-long intervals. Trends use less storage than raw history while preserving statistical information for long-term analysis.

Inputs: itemids (required), limit, time_from, time_till, +1 more. Returns: object

trigger_createTrigger CreateC

Create a new trigger in Zabbix.

Triggers define the conditions under which problems are detected. They use expressions to evaluate item values and determine when to transition from normal to problem state.

Inputs: expression (required), description (required), status, comments, +1 more. Returns: object

trigger_deleteTrigger DeleteA
Destructive

Delete triggers from Zabbix.

Permanently removes one or more triggers. Hosts will no longer generate alerts for these conditions. Historical trigger data and associated problems are typically retained.

Inputs: triggerids (required). Returns: object

trigger_getTrigger GetC
Read-onlyIdempotent

Get triggers from Zabbix.

Triggers are rules that define when a problem occurs based on item values. They evaluate expressions against collected metrics and transition between problem and normal states.

Inputs: limit, offset, output, search, +12 more. Returns: object

trigger_updateTrigger UpdateA

Update an existing trigger in Zabbix.

Modifies properties of an existing trigger. You can change the description, expression, priority, status, or comments. Only specify fields you want to change.

Inputs: triggerid (required), status, comments, priority, +2 more. Returns: object

user_createUser CreateB

Create a new user in Zabbix.

Creates a new user account with specified credentials, role and group membership.

Inputs: passwd (required), usrgrps (required), username (required), name, +2 more. Returns: object

user_deleteUser DeleteA
Destructive

Delete users from Zabbix.

Permanently removes user accounts from the system. The user's access will be immediately revoked. Historical data and previous actions by the user are retained for audit purposes.

Inputs: userids (required). Returns: object

user_getUser GetB
Read-onlyIdempotent

Get users from Zabbix.

Users represent people with access to the Zabbix system. Each user has authentication credentials and permission level determining what they can view and modify.

Inputs: limit, offset, output, search, +5 more. Returns: object

usermacro_createUsermacro CreateB

Create a new host macro in Zabbix.

Host macros define custom variables for specific hosts. They can be referenced in items and triggers using {$MACRO_NAME} syntax, allowing dynamic configuration without editing items.

Inputs: macro (required), value (required), hostid (required), type_, +1 more. Returns: object

usermacro_deleteUsermacro DeleteA
Destructive

Delete host macros from Zabbix.

Permanently removes host-level macro definitions. Items and triggers using this macro will no longer be able to reference it, potentially causing parsing errors.

Inputs: hostmacroids (required). Returns: object

usermacro_getUsermacro GetC
Read-onlyIdempotent

Get user macros from Zabbix.

User macros are variables that can be referenced in items, triggers, and scripts. They allow parameterization of monitoring configurations with custom values.

Inputs: limit, offset, output, search, +9 more. Returns: object

user_updateUser UpdateB

Update an existing user in Zabbix.

Modifies properties of an existing user account. You can change name, surname, password, or role. Only specify the fields you want to change.

Inputs: userid (required), name, passwd, roleid, +2 more. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 53 tools

    Disambiguation4/5

    Most tools map to a distinct resource and action, making core selection straightforward. The main ambiguous pairs are problem_get vs event_get and discoveryrule_get vs drule_get, where descriptions are needed to clarify the boundary.

    Naming Consistency4/5

    The dominant pattern is predictable snake_case resource_action naming, such as host_get, host_create, and trigger_update. Minor deviations include api_version lacking an action verb, drule_get abbreviating discoveryrule_get, and compact compounds like hostgroup and usermacro.

    Tool Count2/5

    With 53 tools, the server is far above the well-scoped 3-15 range and places a heavy burden on tool selection. The broad Zabbix domain justifies many operations, but the set still feels unwieldy and includes many separate read-only getters.

    Completeness3/5

    Core monitoring objects such as hosts, host groups, items, triggers, templates, maintenance, users, and proxies have CRUD coverage, and history/trends/events are readable. However services, SLAs, actions, media types, graphs, discovery rules, and scripts are largely read-only or missing create/update/delete, leaving notable lifecycle gaps.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables users to monitor and query Zabbix infrastructure through natural language by providing access to current problems, active triggers, and system health status via the Zabbix API.
      -
    • A
      license
      C
      quality
      D
      maintenance
      Exposes the complete Zabbix API functionality through the Model Context Protocol, mapping API methods to tools for managing hosts, triggers, and monitoring data. It enables seamless integration and control of Zabbix monitoring environments via natural language interfaces.
      100
      23 PyPI
      MIT
    • A
      license
      B
      quality
      A
      maintenance
      Official initMAX MCP server exposing the Zabbix API to AI assistants for monitoring and managing hosts, problems, templates, dashboards, and multiple Zabbix instances. Includes correlated monitoring views, graphs, anomaly detection, capacity forecasts, and PDF reports, with scoped tokens, read-only modes, OAuth 2.1, and an admin portal. Supports Streamable HTTP, SSE, and stdio.
      84
      251
      AGPL 3.0
    • F
      license
      Not graded
      quality
      C
      maintenance
      Exposes Zabbix monitoring capabilities as callable tools for AI agents and MCP-compatible clients.
      -