helm-mcp
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., "@helm-mcp list all services"
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.
helm-mcp
MCP server exposing the HELM dashboard REST
API (http://localhost:7010 by default) as tools for MCP clients.
Tools
System: helm_health, helm_system_info, helm_export,
helm_scheduler_next_runs.
Services: helm_list_services, helm_get_service, helm_start_service,
helm_stop_service, helm_restart_service, helm_service_logs,
helm_service_metrics, helm_create_service, helm_update_service,
helm_delete_service.
Scripts: helm_list_scripts, helm_get_script, helm_run_script,
helm_run_status.
Service/script tools accept a numeric id or a name.
POST /api/restart-server is intentionally not exposed — the HELM process is
managed by NSSM/systemd and must not be restarted by the agent.
Related MCP server: MCP Tools Hub
Config
Env var | Default | Meaning |
|
| HELM dashboard base URL |
| (empty) | Dashboard PIN (sent as |
Run
uv sync
uv run helm-mcp # stdio transportRegister with Claude Code
claude mcp add helm --scope user -- uv run --directory J:/dagamon_programs/mcp/helm-mcp helm-mcpAdd --env HELM_PIN=<pin> if DASHBOARD_PIN is set in HELM's .env.
Available Tools
35 toolshelm_create_serviceA
Create a service. service is a ServiceCreate object (requires at least
name and type; e.g. {"name": "...", "type": "process", "command": "..."}).
Inputs: service (required).
helm_create_stackB
Create a stack. stack requires "name"; optional "description" and
"tags" (list of strings). Assign services to it by setting their stack_id
via helm_update_service.
Inputs: stack (required).
helm_delete_serviceA
Delete a service by id or name (stops it first if running).
Inputs: name_or_id (required).
helm_delete_stackA
Delete a stack by id or name. Member services survive, detached (their stack_id becomes null); processes are not touched.
Inputs: name_or_id (required).
helm_diagnosticsA
Get host diagnostics: CPU (global + per-core), memory, swap, disks, host uptime, plus a per-managed-process CPU/memory breakdown and HELM's share.
Inputs: no named parameters.
helm_exportA
Export the full HELM config (version, services, scripts) as JSON.
Inputs: no named parameters.
helm_get_docA
Get one HELM documentation/FAQ article by slug. Returns {slug, title, content}.
Inputs: slug (required).
helm_get_scriptA
Get one script by numeric id or by name.
Inputs: name_or_id (required).
helm_get_serviceA
Get one service by numeric id or by name.
Inputs: name_or_id (required).
helm_get_settingB
Get an arbitrary server-stored setting value by key (JSON), or null.
Inputs: key (required).
helm_get_stackB
Get one stack by numeric id or by name.
Inputs: name_or_id (required).
helm_get_themeA
Get the server-stored dashboard theme: {colors: {token: hex, ...}, activePreset: name|null}, or null if never set. Tokens include accent, bg, surface, surface-hover, surface-raised, border, text, text-muted, success, warning, danger.
Inputs: no named parameters.
helm_healthA
Check that the HELM dashboard is reachable. Returns {"status": "ok"}.
Inputs: no named parameters.
helm_list_docsA
List available HELM documentation/FAQ articles as {slug, title}.
Inputs: no named parameters. Returns: result
helm_list_scriptsA
List all scripts.
Inputs: no named parameters. Returns: result
helm_list_servicesA
List all services with their status, pid and metrics.
Inputs: no named parameters. Returns: result
helm_list_stacksA
List all stacks with service_count, running_count and aggregate status (running | partial | stopped).
Inputs: no named parameters. Returns: result
helm_restart_serviceB
Restart a managed service by id or name. Returns {status, pid}.
Inputs: name_or_id (required).
helm_restart_stackC
Restart all services in a stack (stop pass, then start pass).
Inputs: name_or_id (required).
helm_run_scriptA
Run a script by id or name. Returns {run_id, pid, started_at}.
Inputs: name_or_id (required).
helm_run_statusA
Get the status of a script run by run_id (status, exit_code, timestamps).
Inputs: run_id (required).
helm_scheduler_next_runsB
Get next scheduled run times for cron-enabled scripts.
Inputs: no named parameters.
helm_service_logsB
Get recent output log entries for a service (default last 100 lines).
Inputs: name_or_id (required), limit. Returns: result
helm_service_metricsB
Get recent CPU/memory metric snapshots for a service.
Inputs: name_or_id (required). Returns: result
helm_set_settingB
Set an arbitrary server-stored setting. value is any JSON value.
Inputs: key (required), value (required).
helm_set_themeA
Set the dashboard theme. theme is {"colors": {token: "#rrggbb", ...},
"activePreset": name|null}. Connected dashboards pick it up on next load.
Inputs: theme (required).
helm_start_serviceB
Start a managed service by id or name. Returns {status, pid}.
Inputs: name_or_id (required).
helm_start_stackA
Start all services in a stack (intra-stack depends_on order; already running members are skipped). Returns {status, services: [{id, name, outcome, error?}]}; status is "partial" if any member failed.
Inputs: name_or_id (required).
helm_stop_serviceA
Stop a managed service by id or name.
Inputs: name_or_id (required).
helm_stop_stackC
Stop all services in a stack (dependents stop before dependencies). Returns per-service outcomes like helm_start_stack.
Inputs: name_or_id (required).
helm_system_infoA
Get HELM system info: os, runtime version, uptime, service/running counts.
Inputs: no named parameters.
helm_update_applyA
Pull latest, rebuild backend+frontend and restart HELM. Returns once the detached updater has been launched ({status, script, log}).
Inputs: no named parameters.
helm_update_checkA
Check for HELM updates via git: returns {branch, current, latest, behind, ahead, update_available, latest_subject, checked_at}.
Inputs: no named parameters.
helm_update_serviceC
Update a service with a partial ServiceUpdate object of fields to change.
Inputs: updates (required), name_or_id (required).
helm_update_stackC
Update a stack with a partial object: any of name, description, tags.
Inputs: updates (required), name_or_id (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 35 tools
All 35 tools have clearly distinct purposes targeting specific resources and actions. There is no overlap between tools for services, stacks, scripts, themes, settings, docs, or system operations.
Tools follow a consistent 'verb_noun' pattern (e.g., helm_create_service, helm_list_docs), with only a few exceptions like helm_service_logs and helm_scheduler_next_runs that lack a verb prefix. The overwhelming majority are predictable.
35 tools is on the higher side but appropriate given the breadth of the domain: services, stacks, scripts, settings, themes, docs, updates, and diagnostics. Each tool earns its place without redundancy.
The tool surface covers CRUD and lifecycle operations for services and stacks, plus scripts and settings. Minor gaps exist: scripts lack create/update/delete, and settings lack a delete tool. Overall, the core workflows are well-supported.
Maintenance
Related MCP Connectors
MCP server for AI access to Swagger by SmartBear.
An MCP server giving access to Grafana dashboards, data and more.
Linux server administration over OAuth MCP: shell, Docker, systemd, files and monitoring.
Related MCP Servers
- AlicenseNot gradedqualityFmaintenanceAn MCP server that exposes the full API of the Komodo infrastructure management platform. It enables users to manage deployments, stacks, servers, and containers through tools for inspection, configuration, and operational control.MIT
- FlicenseNot gradedqualityDmaintenanceMCP server with a web interface for viewing and calling tools via REST API, enabling easy tool management through a browser.-
- AlicenseNot gradedqualityDmaintenanceMCP server for infrastructure discovery and remote management, enabling SSH command execution, file transfer, log tailing, and machine/service inventory with a companion web dashboard.2MIT
- FlicenseNot gradedqualityDmaintenanceAn MCP server that provides tools for filesystem, database, web, system, and shell operations, along with resources and prompts via SSE transport.-