Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Zabbix-MCP show me all active problems sorted by severity"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A production-oriented Model Context Protocol server for Zabbix 7.4. It lets agent harnesses (Claude Desktop/Code, ChatGPT/Codex, Hermes, OpenClaw, Cursor…) triage problems, query metrics, manage maintenance and build/edit dashboards, with guardrails so an LLM cannot do damage by accident.
Python 3.11+, official mcp SDK (FastMCP), httpx, Streamable HTTP + stdio
Docker image, Traefik path routing (no new hostname needed), interactive Debian installer
Web admin UI for every setting, hot-reloaded
Full API coverage (228 methods / 60 API groups) without 228 tools
Tier | Tools |
Curated |
|
Toolset | one |
Escape hatch |
|
Outputs are trimmed to useful fields, limit defaults to 50, and responses are capped (20 000 chars by default) to protect the context window.
Related MCP server: Zabbix MCP Server
READ_ONLY (default on) hard-blocks every mutating method.
Two-step writes: first call returns a single-use, 5-minute confirm_token bound to method + params (CONFIRM_WRITES=all|destructive|none).
Unknown methods blocked; deny beats allow; users/roles/tokens/settings/modules/script.execute/configuration.import blocked by default.
Secrets never logged (stderr only, masked); JSON-lines audit log; bearer tokens stored as hashes; admin password PBKDF2.
See docs/INSTALL.md. Short version: git clone … && sudo ./install.sh.
See docs/CLIENTS.md for Claude Desktop, Claude Code, ChatGPT desktop/Codex, Hermes, OpenClaw,
Cursor. Skill files that teach agents the workflow live in skills/:
skills/zabbix-mcp (generic/Claude), skills/hermes/zabbix-mcp, skills/openclaw/zabbix-mcp.
uv sync
uv run pytest # 111+ tests, no Zabbix needed (fake API)
uv run ruff check .
ZABBIX_URL=https://zbx ZABBIX_API_TOKEN=… uv run zabbix-mcp check # verify connectivity
ZABBIX_URL=… ZABBIX_API_TOKEN=… uv run python scripts/live_check.py [--write] # exercise tools on a real Zabbixstdio mode (MCP_TRANSPORT=stdio) keeps stdout clean for the protocol.
Unit/integration tests run against a fake Zabbix. CI additionally builds the Docker image and runs a contract test
against a real Zabbix 7.4 stack (scripts/live_check.py --write). Run live_check.py against your own instance after
installing to validate the curated tools end to end.
MIT, see LICENSE.
Acknowledge, comment on, re-rate or close problems. Two-step: first call returns a preview.
Inputs: eventid (required), actions, message, severity, +1 more.
Supported dashboard widget types with their key field names, plus layout rules.
Inputs: no named parameters.
Current unresolved problems (newest first) with host, severity and age.
Inputs: host, group, limit, min_severity, +2 more.
Zabbix audit log entries (who changed what), newest first.
Inputs: user, hours, limit, action.
Status, interfaces, groups, templates, tags and item/problem counts for one host.
Inputs: host (required).
Latest collected values of a host's items, optionally filtered by name/key.
Inputs: host (required), limit, pattern.
Problem events (including resolved ones) in a time window, newest first.
Inputs: host, hours, limit, min_severity.
Zabbix version, HA nodes, proxies, host counts and open problem counts.
Inputs: no named parameters.
SLA/SLI per service for recent periods; without 'sla' lists available SLAs.
Inputs: sla, periods, service.
Triggers (monitored, enabled) with severity and state.
Inputs: host, limit, min_severity, only_problems.
Host groups with host counts.
Inputs: limit, search.
Create and edit Zabbix dashboards. Changes are validated, then two-step confirmed.
Inputs: operation (required), name, page, pages, +6 more.
List, schedule (one-time) or remove maintenance windows. create/delete are two-step.
Inputs: operation (required), name, hosts, groups, +7 more.
History (or trends for windows >48h) of up to 5 matching items, downsampled to min/avg/max buckets.
Inputs: host (required), item (required), minutes, max_points.
Find hosts by name, technical name, IP or DNS.
Inputs: query (required), group, limit.
Zabbix alerting API. Methods: action: create, delete, get, update; mediatype: create, delete, get, update; script: create, delete, execute, get, getscriptsbyevents, getscriptsbyhosts, update; connector: create, delete, get, update. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Zabbix config API. Methods: itemprototype: create, delete, get, update; triggerprototype: create, delete, get, update; graphprototype: create, delete, get, update; hostprototype: create, delete, get, update; discoveryrule: create, delete, get, update; discoveryruleprototype: create, delete, get, update; drule: create, delete, get, update; dhost: get; dservice: get; dcheck: get; httptest: create, delete, get, update; httpstep: get; usermacro: create, createglobal, delete, deleteglobal, get, update, updateglobal; valuemap: create, delete, get, update; regexp: create, delete, get, update; correlation: create, delete, get, update; configuration: export, import, importcompare. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Docs link, read/write class, permission status and usage tips for one API method.
Inputs: method (required).
Zabbix hosts API. Methods: host: create, delete, get, massadd, massremove, massupdate, update; hostgroup: create, delete, get, massadd, massremove, massupdate, propagate, update; hostinterface: create, delete, get, massadd, massremove, replacehostinterfaces, update; template: create, delete, get, massadd, massremove, massupdate, update; templategroup: create, delete, get, massadd, massremove, massupdate, propagate, update; proxy: create, delete, get, update; proxygroup: create, delete, get, update; maintenance: create, delete, get, update. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Zabbix monitoring API. Methods: problem: get; event: acknowledge, get; alert: get; trigger: create, delete, get, update; history: clear, get, push; trend: get; item: create, delete, get, update; graph: create, delete, get, update; graphitem: get; hanode: get; task: create, get; auditlog: get. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Escape hatch: call any 7.4 API method. Same safety rules (read-only, blocklist, confirmation).
Inputs: method (required), params, confirm_token.
Find Zabbix 7.4 API methods by name. Shows which tool runs each and if it is allowed.
Inputs: query (required), limit.
Zabbix services API. Methods: service: create, delete, get, update; sla: create, delete, get, getsli, update. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Zabbix visualization API. Methods: dashboard: create, delete, get, update; templatedashboard: create, delete, get, update; map: create, delete, get, update; iconmap: create, delete, get, update; image: create, delete, get, update; report: create, delete, get, update. Reads return compact defaults; writes are two-step (preview, then repeat with confirm_token). Use zabbix_describe_method for docs and parameters.
Inputs: method (required), params, confirm_token.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 24 tools
The curated convenience tools (get_active_problems, get_problem_events, get_triggers, get_audit_log, manage_dashboard, get_sla_report) overlap directly with the generic grouped tools (zabbix_monitoring, zabbix_visualization, zabbix_services), so an agent must guess whether to use the wrapper or the raw group call. Descriptions hint at the distinction (compact curated reads vs full API surface), but boundaries remain fuzzy for several pairs.
Naming mixes several conventions: verb_noun wrappers (get_problem_events, search_hosts, acknowledge_problem), a zabbix_ namespace prefix (zabbix_monitoring, zabbix_raw_api, zabbix_describe_method), manage_ verbs (manage_dashboard, manage_maintenance_window), and a bare noun (dashboard_widget_catalog). Each name is readable on its own, but there is no single predictable pattern an agent can rely on.
24 tools lands at the borderline-heavy end of the scale, and some of that weight is redundant since the six grouped API tools already subsume much of what the convenience wrappers do. Given Zabbix's very large API surface the count is defensible, but it is not tightly scoped.
Coverage is broad: monitoring, hosts, config, visualization, services, alerting, docs lookup and a raw API escape hatch cover nearly every lifecycle operation, including two-step writes. Minor gaps exist (e.g. user/usergroup/role and token management APIs are not exposed), but agents can work around these via zabbix_raw_api.
Zendesk configuration management for AI agents: propose, review, apply and roll back changes.
The trust harness for AI agents. Set what an agent can do before it acts.
Read-only finance and operations controls for AI agents with evidence and safe next actions.
Provides capabilities that let LLM agents perform a range of infrastructure management tasks.
