health-monitor-mcp

health-monitor-mcp

MCP server health monitoring, uptime history, latency tracking, alert evaluation, and operational reports through natural-language tools.

What This Does

health-monitor-mcp keeps a local registry of MCP servers, performs real MCP handshakes, records health history in SQLite, evaluates alert thresholds, and returns JSON or Markdown evidence suitable for agents and operators.

Supported target transports:

• Streamable HTTP for current remote MCP servers.
• SSE for legacy MCP servers.
• stdio for trusted local executables after explicit opt-in.

Azure DevOps monitoring was retired in v1.1.0. CI-provider integrations are tracked as independent future work.

Quick Start

Run the published package noninteractively with Node.js 24:

npx -y health-monitor-mcp --version

Example MCP client configuration:

{
  "mcpServers": {
    "health-monitor": {
      "command": "npx",
      "args": ["-y", "health-monitor-mcp"]
    }
  }
}

Tools

Tool	Purpose	Typical prompt
register_server	Register an MCP target	Register https://inventory.example.com/mcp as inventory-prod
check_server	Run one live health check	Check inventory-prod now
check_all	Check all matching targets with bounded concurrency	Check all production MCP servers
get_uptime	Return uptime and latency history	Show 24h uptime for inventory-prod
get_dashboard	Return a JSON dashboard	Give me a 24h dashboard
get_report	Return a Markdown report	Generate a 24h health report
list_servers	List registered targets	List monitored servers
unregister_server	Remove a target	Stop monitoring local-debugger
set_alert	Configure thresholds	Alert if inventory-prod exceeds 500ms
get_monitor_stats	Inspect monitor-level activity	How many checks are stored?

Expected configuration mistakes return stable error codes and remediation hints, including SERVER_NOT_FOUND, NO_SERVERS_REGISTERED, STDIO_DISABLED, and STDIO_COMMAND_REJECTED.

Register Targets

Streamable HTTP:

register_server name="inventory-prod" type="http" url="https://inventory.example.com/mcp" tags=["production","inventory"]

Legacy SSE:

register_server name="legacy-search" type="sse" url="https://search.example.com/sse" tags=["legacy"]

Trusted local stdio:

export HEALTH_MONITOR_ALLOW_STDIO=1
export HEALTH_MONITOR_STDIO_ALLOWLIST=npx,node

register_server name="local-debugger" type="stdio" command="npx" args=["-y","mcp-debug-recorder"] tags=["local"]

The command field must contain one executable only. Put package names and flags in args. Remote-safe runtime profiles always disable stdio.

Health Checks and Reports

check_server name="inventory-prod" timeout_ms=5000
check_all timeout_ms=5000 tags=["production"]
get_uptime name="inventory-prod" hours=24
get_dashboard hours=24 include_tool_stats=true
get_report hours=24

HEALTH_MONITOR_MAX_CONCURRENCY limits both scheduled and interactive batch checks. Results preserve registration order even when checks complete out of order.

Alerts

set_alert name="inventory-prod" max_response_time_ms=500 min_uptime_percent=99 consecutive_failures_before_alert=2