Skip to main content
Glama

hcom-mcp

Control-plane MCP server for launching and supervising hcom-managed agents.

Why HTTP, not stdio

MCP servers typically run over stdio — the client spawns one server process per session. For a control-plane server that manages a fleet of agents, that's a problem: ten terminal sessions spawn ten isolated MCP processes with no coordination between them.

hcom-mcp runs as a single persistent HTTP server. Every terminal session, every MCP client, every agent talks to the same endpoint. One process, one source of truth for launches, registries, and lifecycle.

Related MCP server: GRACE Orchestrator MCP

Prerequisites

Install

git clone https://github.com/scchearn/hcom-mcp.git
cd hcom-mcp
npm install
npm run build

Run

Foreground (any OS)

HCOM_MCP_PORT=3111 node dist/index.js

Listens on http://127.0.0.1:3111/mcp. Only localhost connections are accepted.

As a daemon

OS

Mechanism

Setup

Linux

systemd user service

cp docs/systemd/hcom-mcp.service ~/.config/systemd/user/ && systemctl --user enable --now hcom-mcp

macOS

launchd

cp docs/launchd/com.scchearn.hcom-mcp.plist ~/Library/LaunchAgents/ && launchctl load ~/Library/LaunchAgents/com.scchearn.hcom-mcp.plist

Windows

Task Scheduler or nssm

See docs/windows/hcom-mcp-task.xml for Task Scheduler, or the nssm commands in the file comments

Adjust the ExecStart/ProgramArguments path in the config to match where you cloned the repo.

MCP client configuration

Point any HTTP-capable MCP client at http://127.0.0.1:3111/mcp. Example for Claude Desktop:

{
  "mcpServers": {
    "hcom-mcp": {
      "type": "http",
      "url": "http://127.0.0.1:3111/mcp"
    }
  }
}

Configuration

  • HCOM_MCP_PORT — HTTP port (default: 3111)

  • ~/.hcom/mcp/config.json — presets, topologies, model catalogs, rescue allowlist

  • ~/.hcom/mcp/registry.json — managed agent registry

  • .hcom-mcp.json — optional workspace overlay (see .hcom-mcp.example.json)

Rescue allowlist

unblock and spawn_and_verify only inject input into a blocked agent when the pending launch_blocked detail matches a pattern in rescueAllowlist. Defaults cover the known rescuable dialogs (workspace trust, permission mode, model/provider picker); add patterns for new dialogs without a code release:

{
  "rescueAllowlist": {
    "enabled": true,
    "patterns": ["trust this folder", "permission mode"]
  }
}

Workspace overlays extend the global patterns; they cannot remove the built-in defaults.

Skills

Install with:

npx skills add scchearn/hcom-mcp

Operating guidance ships in skills/:

  • skills/using-hcom/SKILL.md — messaging, threads, transcripts, event watching

  • skills/hcom-agent-messaging/SKILL.md — agent-to-agent messaging patterns and gotchas

Tools

Tool

Purpose

launch

Launch a headless agent (preset or bare harness+model)

spawn_and_verify

Launch + gate on readiness, with optional guarded rescue of blocked agents

launch_topology

Launch multiple agents from a topology preset (optional verify gate)

adopt

Adopt an external hcom agent into managed lifecycle

stop / kill

Stop or kill managed agents (one or more names, or a tag)

unblock

Guarded PTY rescue for a blocked agent (dry-run by default, config allowlist)

watch_agents

Supervise owned agents: poll snapshot with derived flags, or subscribe to life/blocked events

resume / fork

Resume a stopped agent or fork a session, registering ownership with a resumedFrom link

send

Send an hcom message to one or more agents (request/inform/ack, optional reply_to)

list_managed

List agents managed by this server

list_all

List all live hcom agents

list_models

List available models per harness (claude full IDs pass through unverified)

list_presets

List configured agent presets

list_topologies

List configured topology presets

inspect

Inspect an agent's status, transcript, events, or terminal

transcript

Read agent transcripts, transcript search results, or transcript timeline

continue_from

Get handoff context from a live or stopped agent

prune

Remove stale registry records

thread_seed

Create a workflow thread

thread_inspect

Query thread events

status

Server health and orientation (includes hcom status --json health, config/registry paths)

License

MIT

Related MCP Connectors

Related MCP Servers

  • A
    license
    Not graded
    quality
    B
    maintenance
    MCP server that enables AI agents to run a deterministic orchestration loop with decomposition, subagent execution, and review feedback across multiple LLM backends.
    60
    MIT
  • F
    license
    A
    quality
    C
    maintenance
    MCP server for controlling a Hermes agent deployment via its HTTP API, enabling health checks, log retrieval, task management, and deployment configuration from MCP clients.
    17
    -
  • A
    license
    Not graded
    quality
    A
    maintenance
    A self-hosted MCP control plane and protocol gateway that aggregates and manages multiple MCP servers, exposing them via stable, authenticated MCP endpoints for any client.
    0
    1
    MIT