Ctrlk
Our Services
For the complete documentation index, see llms.txt. This page is also available as Markdown.

vMonitor MCP Tools

vMonitor MCP Server provides tools to manage and automate the vMonitor Platform — GreenNode's observability service: dashboards & widgets, metric queries, alarms, infrastructure hosts, logs, notifications, quota & usage, and synthetic uptime monitors. The server fronts five vMonitor APIs (metric/dashboard, Log, notification, quota-usage, synthetic/uptime) behind a single IAM authentication — requests are routed to the right API automatically. Every operation is exposed as a tool with argument-based input — no resource URIs are used.

Tool groups

Group
Purpose

Dashboards

List / inspect / create / clone / update / delete dashboards.

Widgets, variables & views

Add / edit / move / resize widgets; manage shared variables and saved views.

Metric query

Time-series and single-value statistics — the data behind every chart.

Metric catalogue & units

Metric names, dimensions, values, units and per-user unit overrides.

Infrastructure hosts

Agent-based hosts and product hosts (vServer, vStorage, vDB, vLB, vBackup, vBandwidth, vAS) + their metric snapshots.

Alarms

Metric / log / change-detection alarms: create / update / delete, histories and statuses.

Integrations & metric API keys

Install / uninstall metric-source integrations; issue / revoke metric API keys.

Logs — projects & search

Log projects, certificates, field mappings, log search and export.

Logs — archives & refills

Export destinations (archives) and re-ingest jobs (refills).

Logs — pipelines & processors

Processing pipelines, processor groups, processors and libraries.

Logs — resource mappings

Resource → project log mappings (vCDN, vDB, vLB, vStorage, vStorage bucket).

Notifications

OTP-verified channels: Email, SMS, Slack, Webhook, Telegram, Teams.

Quota & usage

Usage reads, tier / package catalog, price quotes.

Quota orders

Buy / resize / delete quotas — spends money; every order has a zero-cost pre-flight quote.

Synthetic

Uptime monitors and probing locations.

Feature guide

Step-by-step guide for each composite feature.

Conventions

  • verb_noun naming, one handler per feature area.

  • Access is noted per tool:

    • read — always available. The default read-only mode registers 110 tools.

    • write — requires the --allow-write flag. These tools are not registered without it, so the agent cannot see them.

    • destructive — also requires --allow-write; the tool deletes data irreversibly or spends money. Clients warn before the call.

  • Global service: vMonitor has no regions — no tool takes a region parameter, and GRN_DEFAULT_REGION is ignored.

  • Annotations: every tool declares readOnlyHint / destructiveHint so clients can auto-approve reads and warn before destructive calls.

  • Structured JSON output with outputSchema + structuredContent — clients parse it directly.

  • Paging: list tools take 1-based page / size (omit page to fetch everything where noted); log list endpoints use a content-based paging envelope.

  • Money-spending: quota-order tools (create_log_project, resize_*) spend money — always call get_creation_price / get_resize_price first (free, and it validates the payload).

Supported Tools

Dashboards

Tool
Access
Description

list_dashboards

read

List dashboards; optional searching_text / searching_field filter, 1-based page / size paging (omit page for all).

get_dashboard

read

Single dashboard by ID (includes widget count).

get_dashboard_by_name

read

Single dashboard by exact name.

create_dashboard

write

Create an empty dashboard (only name required).

create_dashboard_clone

write

Clone a dashboard into a new one.

update_dashboard

write

Update general settings (dark mode, refresh, time range, selected view).

update_dashboard_name

write

Rename a dashboard.

update_dashboard_favorite

write

Mark / unmark favorite.

delete_dashboard

destructive

Delete a dashboard (irreversible).

Widgets, variables & views

Tool
Access
Description

list_dashboard_variables

read

Shared query variables of a dashboard.

get_dashboard_variable

read

One shared variable.

update_dashboard_variables

write

Replace a dashboard's variable list (whole-list replace).

list_dashboard_views

read

Saved query / filter / time-range presets.

get_dashboard_view

read

One saved view.

create_dashboard_view

write

Save the current dashboard state as a named view.

update_dashboard_view

write

Update a saved view's stored state.

delete_dashboard_view

destructive

Delete a saved view (irreversible).

list_widgets

read

Dashboard widgets plus the metric query behind each — replay via get_statistics_v2 with no dimension discovery.

get_widget

read

One widget (chart config + graph specs).

create_widget

write

Add a widget (v2 graphs map; omit layout to auto-place on the 10-column grid).

update_widget

write

Edit widget content (v1 metricGraphs / logGraphs arrays).

update_widget_v2

write

Edit widget content (v2 graphs map).

update_widget_layout

write

Move / resize a widget and adjust its time window.

delete_widget

destructive

Delete a widget (irreversible).

Metric query

Tool
Access
Description

get_statistics

read

Time-series data (filter by dimensions, group_by, window) — the data behind a chart.

get_statistics_synthetic

read

Single aggregated value (number / single-stat charts).

get_statistics_v2

read

Typed statistic query (type = SIMPLE / CUSTOM + data body) — replays a widget's query as-is.

Metric catalogue & units

Tool
Access
Description

get_metric_names

read

Metric catalogue — the starting point for picking a metric.

list_metric_dimension_names

read

Every dimension key known across metrics.

list_metric_dimension_values

read

Observed values of one dimension (e.g. hosts for host).

get_metric_dimensions

read

A metric's dimensions + each dimension's observed values.

list_metric_units

read

Units assignable to a metric.

list_metric_unit_mappings

read

Metric → unit mappings shown in info panels.

create_metric_unit_mapping

write

Override a metric's display unit for the current user.

delete_metric_unit_mapping

destructive

Reset the display unit (remove the override).

Infrastructure hosts

Agent-based hosts (Metric Agent servers) plus product resources monitored as hosts. The <type> families expand to one tool per product type — vserver, vstorage, vdb, vlb, vbackup, vbandwidth, vas (vDB Kafka has a list tool only). Host-listing tools always send page / size and accept an optional name filter.

Tool
Access
Description

list_hosts

read

Agent-based infrastructure hosts.

get_host

read

One agent-based host by ID.

get_host_metrics

read

Host's current metric snapshot (status + CPU / load / memory).

update_host_enabled / update_host_disabled

write

Resume / pause monitoring of an agent-based host (agent stays installed).

delete_host

destructive

Remove an agent-based host from the Infrastructure list (irreversible).

list_vserver_hosts / list_vstorage_hosts / list_vdb_hosts / list_vdb_kafka_hosts / list_vlb_hosts / list_vbackup_hosts / list_vbandwidth_hosts / list_vas_hosts

read

List product resources monitored as hosts.

get_<type>_host_metrics

read

Metric snapshot for a product host.

update_<type>_host

write

Enable / disable monitoring for a product host (body {enabled}).

delete_<type>_host

destructive

Remove a product host (irreversible).

Alarms

Metric, synthetic, log and change-detection alarms. Create / update bodies are flat, fully-typed DTOs — name is the only strictly required field.

Tool
Access
Description

list_alarms

read

List alarms (filter by name / severity / status / type).

get_alarm

read

One alarm by ID (with type-specific config).

get_metric_alarm_definition

read

Metric alarm's upstream evaluator definition.

list_metric_alarm_histories

read

Metric alarm history (id = a sub-alarm alarms[].id from the definition).

get_synthetic_alarm_definition

read

Synthetic metric alarm's definition.

list_synthetic_alarm_histories

read

Synthetic alarm history (id = sub-alarm from the definition).

list_log_alarm_histories

read

Log alarm history.

get_log_alarm_status

read

Log alarm's current status.

get_change_alarm

read

Change-detection alarm's definition (requires a time window).

list_change_alarm_histories

read

Change alarm history (requires a time window).

create_metric_alarm / update_metric_alarm

write

Create / edit a metric alarm (severity / condition case-insensitive).

delete_metric_alarm

destructive

Delete a metric alarm (irreversible).

delete_metric_sub_alarm

destructive

Delete a composite metric alarm's sub-alarm.

create_log_alarm / update_log_alarm

write

Create / edit a log alarm.

delete_log_alarm

destructive

Delete a log alarm (irreversible).

create_change_alarm / update_change_alarm

write

Create / edit a change-detection alarm.

delete_change_alarm

destructive

Delete a change alarm (irreversible).

delete_change_alarm_history

destructive

Clear change alarm history (irreversible).

Integrations & metric API keys

Tool
Access
Description

list_integrations

read

Installable metric-source apps.

get_integration

read

One integration.

update_integration_installed / update_integration_uninstalled

write

Install / uninstall an integration.

delete_integration

destructive

Delete an integration (irreversible).

list_metric_api_keys

read

List metric API keys.

create_metric_api_key

write

Issue a new metric API key.

delete_metric_api_key

destructive

Revoke a metric API key (irreversible).

Logs — projects, search & export

search_logs / search_logs_default take a structured {type,value} DSL (match / range / exists / bool; Elasticsearch shorthands are translated).

Tool
Access
Description

list_projects

read

Log projects (Log API).

get_project

read

One log project.

get_project_mappings

read

A project's field mappings.

get_project_log_data_exists

read

Whether a log project has ingested data.

search_logs / search_logs_default

read

Query a project's data with the structured DSL.

get_log_export

read

Track an asynchronous log export.

get_project_certificate_download

read

Download a project certificate (base64 ZIP; cert_id from list_projects → certInfos[]).

update_project / update_project_mappings

write

Edit a project's settings / field mappings.

create_project_certificate

write

Issue a project client certificate.

delete_project_certificate

destructive

Revoke a project client certificate.

create_log_export

write

Prepare an asynchronous log export.

Logs — archives & refills

Tool
Access
Description

list_archives / get_archive

read

Log archives (export destinations).

validate_archive_connection

read

Test archive storage connectivity.

create_archive / update_archive

write

Create / edit a log archive.

delete_archive

destructive

Delete a log archive.

list_refills / get_refill

read

Log refill (re-ingest) jobs.

validate_refill_connection

read

Test refill storage connectivity.

create_refill / create_refill_from_archive

write

Create a refill job.

delete_refill

destructive

Delete a refill job.

Logs — pipelines & processors

Tool
Access
Description

list_pipelines / get_pipeline

read

Log processing pipelines.

create_pipeline / update_pipeline

write

Create / edit a pipeline.

delete_pipeline

destructive

Delete a pipeline.

get_processor_group

read

One processor group.

list_processor_group_libraries

read

Processor group libraries.

list_date_formats

read

Date-format helpers.

validate_grok_parser

read

Validate a grok parser.

create_processor_group / update_processor_group / update_processor_order / create_processor_group_library

write

Manage processor groups and their order.

delete_processor_group

destructive

Delete a processor group.

create_processor / update_processor

write

Manage processors.

delete_processor

destructive

Delete a processor.

Logs — resource mappings

The <product> placeholder expands to vcdn, vdb, vlb, vstorage (plus a dedicated vStorage bucket tool).

Tool
Access
Description

list_<product>_log_mappings (+ list_vstorage_bucket_log_mappings)

read

Resource → project log mappings.

list_vcdn_log_mapping_types / list_vstorage_log_mapping_regions

read

Mapping type / region lookups.

update_<product>_log_mapping[_enabled|_disabled]

write

Edit / enable / disable a resource log mapping.

update_vstorage_bucket_log_mapping

write

Set a vStorage bucket's log mapping.

Notifications

Channel creation is OTP-verified: create_notification_otp → validate_notification_otp → create_notification. Channels cover Email, SMS, Slack, Webhook, Telegram, Teams.

Tool
Access
Description

list_notification_types

read

Channel types.

list_notifications

read

Notification channels.

get_notification_otp_info

read

Pending OTP info.

create_notification_otp

write

Send an OTP to a channel address.

validate_notification_otp

write

Validate the OTP.

create_notification / update_notification

write

Create / edit an (OTP-verified) channel.

delete_notification

destructive

Delete a notification channel (irreversible).

Quota & usage (reads — free)

Tool
Access
Description

get_quota_usage / get_log_usage / get_composite_usage

read

Usage per category / per log project / combined.

get_current_quota / list_log_quotas / get_log_quota / get_quota_detail

read

Current active quota and its detail.

get_billing_settings / list_trash_quotas / get_convert_result

read

Billing settings / trashed quotas / conversion result.

list_tiers / get_tier / get_tier_description

read

Quota tier catalog (metric / synthetic / log).

list_packages / get_package / get_package_detail / get_package_description / get_package_description_detail

read

Purchasable package catalog.

list_quota_classes / list_quota_class_packages

read

v2 quota classes and their packages.

get_creation_price / get_resize_price / get_recovery_price / get_renewal_price

read

Price quotes (no order placed).

Quota orders (money-spending; --allow-write only)

Every order has a zero-cost pre-flight via get_creation_price / get_resize_price.

Tool
Access
Description

create_log_project

write

Buy a new log project — the log quota order creates the project (spends money).

resize_log_project

destructive

Grow quota / upgrade Basic → Pro (spends money, irreversible).

delete_log_project

destructive

Delete a project, its quota and stored logs (irreversible).

resize_metric_quota

destructive

Resize the account's single metric quota (spends money, irreversible).

resize_sms_quota / resize_email_quota

destructive

Swap SMS / email notification quota to another package (spends money, irreversible).

Renew and recover-from-trash remain quote-only. Orders respond with order_id, amount and payment_url — a non-empty payment_url means pending (the quota changes only after the user pays via that link); pay: true charges the account directly.

Synthetic (uptime monitors & locations)

Tool
Access
Description

list_uptimes / get_uptime / get_uptime_config / validate_uptime

read

Uptime monitors + probe preview.

create_uptime / update_uptime / update_uptime_status

write

Create / edit / toggle a monitor.

delete_uptime

destructive

Delete a monitor (irreversible).

list_locations / get_location

read

Synthetic probing locations.

create_location / update_location

write

Create / edit a private probing location.

delete_location

destructive

Delete a probing location (irreversible).

Feature guide

Tool
Access
Description

get_feature_guide

read

Step-by-step guide for a composite feature: build_dashboard, query_metrics, create_metric_alarm, monitor_infrastructure, edit_metric_unit, manage_log_projects, manage_integrations, create_notification_channel, view_quota_usage, create_uptime_monitor.

Prompts (11)

The server ships 11 prompts (Vietnamese) — step-by-step feature guides, always available (no --allow-write needed). Load them from the client's prompt list (e.g. the prompt picker in Claude Code), or let the agent call the get_feature_guide tool with the same feature key.

Prompt
Purpose

vmonitor_getting_started

Onboarding: concepts, auth setup, no-region model, tool routing, feature map.

vmonitor_build_dashboard

Dashboards + variables + views + widgets (incl. the graphs / auto-layout shape).

vmonitor_query_metrics

Query metrics / plot data, search & export logs (incl. the log search DSL).

vmonitor_create_metric_alarm

Create an alarm (metric / log / change-detection): source → threshold → notification → confirm gate.

vmonitor_monitor_infrastructure

Explore infrastructure hosts and their metrics.

vmonitor_edit_metric_unit

Override a metric's display unit.

vmonitor_manage_log_projects

Manage log projects, mappings, certificates.

vmonitor_manage_integrations

Install / uninstall metric-source integrations.

vmonitor_create_notification_channel

OTP-verified notification-channel creation.

vmonitor_view_quota_usage

Read quota usage and prices, then buy / resize / delete a quota.

vmonitor_create_uptime_monitor

Create a synthetic uptime monitor + probing location.

Key workflows

Read a resource's metrics off its default dashboard

Every GreenNode resource owns an auto-generated system dashboard, and each widget stores the exact query the console plots (metric name, statistic, grouping, full dimensions string carrying the resource_id). So "how is this server doing?" needs no metric-catalogue walk:

Reusing the widget's period matches the web chart. Widgets with log_graph_count > 0 plot log data — query those with search_logs.

Buy or resize a quota (money-spending, --allow-write)

packageId lives on a quota class's retention entries, along with the bounds the quantity must respect. The quote call both prices and validates the payload:

Leave redirectUrl alone — each order tool fills in the vMonitor console's quota page automatically. Resizing follows the same shape (get_resize_price → resize_log_project; resize_metric_quota uses host counts and minResource / maxResource / step). SMS / email are fixed bundles from list_packages category=sms|email — no quantity.

Usage notes

  • Argument-based input, no resource URIs. Every operation is a tool.

  • Read-only by default (110 tools). Writes need --allow-write — see Configure Local MCP. On a hosted endpoint the access level is fixed by whoever deployed it.

  • Quote before ordering, inspect before deleting. Quota orders always have a free pre-flight quote; before irreversible deletes, read the resource's detail / history tools first.

  • Every infrastructure host owns an auto-generated, read-only default dashboard — the fastest path to a resource's metrics.

PreviousConfigure Remote MCP
NextFeatures of vMonitor Platform

Last updated