whats-running-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| WR_AGENT_PATTERNS | No | Comma-separated substrings that mark a process as an 'agent' in agent_sessions | claude,codex,aider,cursor-agent,copilot,goose |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| agent_sessionsA | List AI-agent processes actually running right now (Claude Code, Codex, Aider, ...). Separates terminal-attached sessions (a human can see them) from detached/orphan ones. Reads ps/lsof directly — never transcripts or docs. Patterns configurable via WR_AGENT_PATTERNS. |
| listening_portsA | All TCP ports currently in LISTEN state with owning process and pid (lsof). The definitive answer to 'is service X actually up?' — nothing hidden, nothing assumed. |
| daemonsA | Persistent services actually loaded right now: launchctl (macOS, non-Apple) or systemd user services (Linux). Optional case-insensitive substring filter on the label. |
| system_statsA | Load average, uptime, and root-disk free space. |
| whats_runningA | Full live snapshot in one call: agent sessions, TCP listeners, loaded daemons, system stats. Use at session start to ground yourself in what is ACTUALLY live instead of stale memory. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool targets a distinct aspect of system state: daemons for persistent services, system_stats for machine health, listening_ports for network listeners, and agent_sessions for AI processes. The aggregate tool whats_running is explicitly designed to combine these, so there is no ambiguity between specific tools.
All tool names use lowercase snake_case and are noun phrases describing their data (daemons, system_stats, listening_ports, agent_sessions). The one outlier is whats_running, which is a phrase rather than a simple noun, but it still follows the same formatting and is easily distinguishable.
With five tools, the server is well-scoped for its monitoring purpose. Each tool covers a specific area, and the aggregate whats_running tool adds value by providing a single-call snapshot without bloating the count.
The server covers the key areas of system state: daemons, system stats, listening ports, and AI agent sessions, with a combined snapshot. Minor gaps exist, such as a general process list or memory usage details, but these are not core to the defined scope and can be worked around.