Remote Agent Ops Console

What it does

Gives a safe workflow for operating a self-hosted automation agent running on a remote Linux/macOS server via SSH. Starts with a read-only health snapshot (logs, cron status, disk, running processes), then proposes fixes or changes that always require explicit confirmation before touching the live server.

When to use it

• A self-hosted agent or automation server stops responding or behaves unexpectedly
• You need to review or update cron jobs, skills, or config on a remote automation box
• You run client automations (including GoHighLevel integrations) on your own infrastructure and need safe remote maintenance

Steps

1. Connect via SSH using stored credentials/keys, never typed secrets in chat
2. Pull a read-only snapshot: service status, recent logs, cron list, disk/memory usage
3. Mask any sensitive values before displaying snapshot output
4. Diagnose the issue from the snapshot and draft a proposed fix
5. Present the exact commands or config changes for approval before running anything that writes to the server
6. Apply only approved changes, then re-snapshot to confirm the fix worked

The skill

---
name: remote-agent-ops-console
description: Connect over SSH to a self-hosted AI agent server to check health, review cron jobs and skills, and fix issues with read-only snapshots first.
---

# Remote Agent Ops Console

## 1. Overview
Gives a safe workflow for operating a self-hosted automation agent running on a remote Linux/macOS server via SSH. Starts with a read-only health snapshot (logs, cron status, disk, running processes), then proposes fixes or changes that always require explicit confirmation before touching the live server.

## 2. When to Use
- A self-hosted agent or automation server stops responding or behaves unexpectedly
- You need to review or update cron jobs, skills, or config on a remote automation box
- You run client automations (including GoHighLevel integrations) on your own infrastructure and need safe remote maintenance

## 3. Steps
1. Connect via SSH using stored credentials/keys, never typed secrets in chat
2. Pull a read-only snapshot: service status, recent logs, cron list, disk/memory usage
3. Mask any sensitive values before displaying snapshot output
4. Diagnose the issue from the snapshot and draft a proposed fix
5. Present the exact commands or config changes for approval before running anything that writes to the server
6. Apply only approved changes, then re-snapshot to confirm the fix worked

## 4. GoHighLevel Angle
Apply this inside GoHighLevel workflows (funnels, automations, or client reporting) wherever the same task recurs for an agency or client account.

## 5. Safety Rules
- Always produce a draft first; never send, publish, or spend money without explicit human approval.
- Never invent data, metrics, or results — use only what is confirmed or clearly label estimates.
- Never paste API keys, tokens, or passwords into chat; reference secrets by name from a secure store.

For Claude Code, save it as ~/.claude/skills/remote-agent-ops-console/SKILL.md. In Cursor, ChatGPT or another assistant, paste it as project instructions.