agentleFS
seayniclabsseayniclabs/homelab-infra-mcp/skills/getting-started/SKILL.md

getting-started

Skill0 starsChanged 18 days agoMITView on GitHub

Set up and check the Homelab MCP server. Covers which modules are on, the settings each module needs, how to switch a module on or off, the first checks, how approvals work, and how to read an error. Use when the user installs Homelab, asks what it can do, gets a homelab.<module>.<code> error, or wants a module turned on or off.

  • Reads credentials

32 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 2 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 32 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Call homelab_modules first.

    line 15
  2. Distinctive:

    Every value above is a placeholder.

    line 63
  3. Distinctive:

    Run these in order. All are read only.

    line 73
  4. Very specific:

    Every change tool accepts dry_run set to true.

    line 87
  5. Very specific:

    Every decision goes into the approval log.

    line 88
  6. Very specific:

    Run homelab-infra --doctor in a terminal.

    line 96
  7. Very specific:

    Every error reads homelab.<module>.<code>: <message> Fix: <hint>.

    line 100

“Common” is measured: kept by at least 6 in 10 of the 32 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Get started with Homelab

Homelab connects an assistant to Nginx Proxy Manager, Docker, and Cloudflare DNS, and to four optional services. The first five minutes take four steps: see what is on, add the missing settings, run the first checks, and learn how approvals work.

1. See what is on

Call homelab_modules first. It returns each module with whether it is enabled, which settings are missing, and what each tool does. Report the missing settings by name. Never guess a value.

Default state:

ModuleDefaultTools
proxyonproxy_hosts, proxy_redirections, proxy_certificates, proxy_access_lists
dockerondocker_containers, docker_images, docker_inventory
dnsondns_zones, dns_records
home_assistantoffha_entities, ha_automations, ha_system
uptimeoffuptime_monitors
tunnelofftunnel_routes
dns_filteroffdns_filter_status, dns_filter_queries, dns_filter_rules
ssooffsso_applications

The core tools are always on: homelab_modules, homelab_health, expose_service, teardown_service, service_status, infrastructure_overview, diagnose_service.

2. Add the settings

Set each value as an environment variable in the MCP client's env block, or in ~/.config/homelab/config.toml. Set HOMELAB_CONFIG to use a different file. Environment variables win over the file. Restart the MCP client after you change a setting.

Each module needs only its own settings:

ModuleVariables
proxyHOMELAB_PROXY_URL (default http://localhost:81), HOMELAB_PROXY_EMAIL, HOMELAB_PROXY_PASSWORD
dockerHOMELAB_DOCKER_HOST (default unix:///var/run/docker.sock); for Portainer, HOMELAB_DOCKER_PORTAINER_URL, HOMELAB_DOCKER_PORTAINER_TOKEN, HOMELAB_DOCKER_PORTAINER_ENDPOINT_ID
dnsHOMELAB_DNS_API_TOKEN, HOMELAB_DNS_ZONE_ID
home_assistantHOMELAB_HOME_ASSISTANT_URL, HOMELAB_HOME_ASSISTANT_TOKEN
uptimeHOMELAB_UPTIME_URL, HOMELAB_UPTIME_API_KEY; HOMELAB_UPTIME_USERNAME and HOMELAB_UPTIME_PASSWORD only for pause and resume
tunnelHOMELAB_TUNNEL_API_TOKEN, HOMELAB_TUNNEL_ACCOUNT_ID, HOMELAB_TUNNEL_TUNNEL_ID
dns_filterHOMELAB_DNS_FILTER_URL, HOMELAB_DNS_FILTER_USERNAME, HOMELAB_DNS_FILTER_PASSWORD
ssoHOMELAB_SSO_URL, HOMELAB_SSO_TOKEN, HOMELAB_SSO_FLOW_SLUG

Any secret can come from a file instead: set HOMELAB_<MODULE>_<KEY>_FILE to the file path, for example HOMELAB_PROXY_PASSWORD_FILE=/run/secrets/npm_password. Secrets are *_PASSWORD, *_TOKEN, *_API_KEY and *_API_TOKEN values.

Config file example:

[proxy]
email = "admin@example.com"
password = "change-me"

[dns]
api_token = "your-cloudflare-token"
zone_id = "your-zone-id"

Every value above is a placeholder. Use the user's real values only in their own config.

3. Switch a module on or off

Set HOMELAB_<MODULE>_ENABLED to true or false. Examples: HOMELAB_UPTIME_ENABLED=true, HOMELAB_DNS_FILTER_ENABLED=true. Restart the MCP client, then call homelab_modules to confirm.

New modules ship off, so an update does not add tools until the user turns them on.

4. Run the first checks

Run these in order. All are read only.

  1. homelab_modules: confirm which modules are on and which settings are missing.
  2. homelab_health: connectivity and version for each enabled module.
  3. proxy_hosts with action set to list: the proxy works and lists its hosts.
  4. docker_containers with action set to list: the Docker socket or Portainer answers.
  5. dns_zones with action set to list: the Cloudflare token reads its zones.

5. How approvals work

Homelab changes nothing on a read or a dry run. Every change asks the person first:

  • The person approves each change in the MCP client, which shows the exact change: the domain, the container, the DNS record.
  • The assistant cannot approve a change for the person. No argument sets approval.
  • Every change tool accepts dry_run set to true. It shows the plan and changes nothing.
  • Every decision goes into the approval log. The default path is ~/.local/state/homelab/approvals.jsonl, or $XDG_STATE_HOME/homelab/approvals.jsonl when XDG_STATE_HOME is set. HOMELAB_APPROVAL_LOG overrides the path.

If the MCP client cannot show an approval prompt, the change returns approval_unavailable. The message names the setting that pre-approves the module, HOMELAB_<MODULE>_ALLOW_WRITES=true, or it says to leave changes alone. Explain both options to the user. Do not set HOMELAB_<MODULE>_ALLOW_WRITES yourself. Pre-approval applies to every change in that module, so it is the person's decision.

HOMELAB_READ_ONLY=true refuses every change, including pre-approved ones.

6. Diagnose a setup

Run homelab-infra --doctor in a terminal. It checks each enabled module's settings and reachability, the Docker socket access and the certificate tool. It prints a FIX line for each problem and exits with status 1 when something needs fixing.

Errors

Every error reads homelab.<module>.<code>: <message> Fix: <hint>. The module tells you where to look. Pass the Fix: line to the user as written. If the error names a setting, check that the setting is set in the environment or the config file, then restart the MCP client.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.