agentleFS
Lukk17Lukk17/agent-standards/.agents/skills/samsung-tv/SKILL.md

samsung-tv

Skill1 starsChanged 3 days agoMITView on GitHub

Controls a Samsung Tizen smart TV from 2016 on through Home Assistant, with read-only checks against its local REST API: launching apps with a Home Assistant REST command, confirming app IDs, turning it on with a guarded power key, voice assistant scripts and casting a dashboard. Use when the user says "open Netflix on the TV", "find the Disney+ app ID", "the TV does not turn on from Home Assistant" or "show a dashboard on the TV". Not for Home Assistant YAML style in general, use `home-assistant`.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never print the websocket token from the config entry, and never copy it out of Home Assistant.

    line 60

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “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

Samsung TV

How an agent drives a Samsung Tizen TV from Home Assistant and checks it directly without changing anything on it. The TV answers a REST API on port 8001 with no login, which is enough to read its state and to confirm whether an app is installed. Every command that changes something goes through Home Assistant, and every change there is asked of the owner first: an app launch through a rest_command that posts to port 8001, and a key or power through the samsungtv integration, which holds the token for the secure websocket on port 8002.

Written for Tizen TVs from 2016 on that use the websocket API, and Home Assistant 2026.10 and newer, where the samsungtv integration no longer sends Wake on LAN itself. A fact marked VERIFIED was measured on a 2024 QN85D in the Polish region, the launch and power facts on 2026-11-06 with Home Assistant 2026.10.4, or read in the source or vendor page that the reference file names. INFERRED carries a confidence and a reason. Everything else is UNVERIFIED and stays marked until it is measured.

You are about toOpen
Read the TV's device info, network type, MAC, Frame support or power state from port 8001, or what the launch POST answerstv-rest-api.md
Find, confirm or choose an app ID, or explain why Home Assistant lists no appsapp-ids.md
Launch an app, change a channel or send a remote key from Home Assistant, explain why play_media opens no app, or tell the two media players aparthome-assistant-control.md
Turn the TV on from Home Assistant, explain why the magic packet does not wake it, or why the entity has no turn onpower-on.md
Cast a Home Assistant dashboard to the TV, or put custom content in Ambient Modecast-and-ambient.md
Write the script a voice phrase runs to open a TV app, or give it several phrases on Alexa, Google Assistant or Assistvoice-assistants.md

Facts about one setup

The TV's address, its MAC address, its entity IDs, the Home Assistant address and the token of a real setup belong in the owner's own notes, for example MACHINE.md next to the global AGENTS.md, never in this skill and never in a shared repository. Every example uses placeholders in angle brackets: <tv-ip>, <tv-mac>, <tv-entity> for the samsungtv media player, <tv-remote-entity> for its remote, <cast-entity> for the Google Cast one, <app-id>, <dashboard-path>, <view-path>.


When to activate

  • Launching an app on the TV from Home Assistant, a script or an automation
  • Finding or confirming the app ID of Netflix, YouTube, Disney+, Spotify or any other TV app
  • Turning the TV on or off from Home Assistant, or a TV entity that offers no turn on
  • Writing a voice assistant script that opens a TV app
  • Casting a Home Assistant dashboard view to the TV
  • Reading the TV's network type, MAC address or Frame support without touching its settings
  • The user says "which Netflix ID works on my TV", "the app script runs but nothing opens", "Wake on LAN stopped working after the update" or "can Ambient Mode show my dashboard"

When not to activate

  • Home Assistant YAML style, automation modes, secrets and backups in general, use home-assistant
  • The Docker VM, its Compose stack and its services, use homelab-services
  • The router, its DHCP leases and its Wi-Fi, use asus-router
  • SmartThings appliances such as a washer or a fridge, which this skill does not cover
  • Judging whether the TV's open ports or the token are acceptable, use the security-auditor agent

Hard rules

  1. The agent only reads the TV directly: GET on http://<tv-ip>:8001/api/v2/ and http://<tv-ip>:8001/api/v2/applications/<app-id>. It never opens the websocket, sends a key, launches an app or sends any other method to the TV itself. The launch POST is sent by Home Assistant, never by the agent.
  2. Ask before every change in Home Assistant: an automation, a script, an integration option, a restart, and every service call that acts on the TV. Say what changes, the worst case and how to undo it, then wait for a yes. One yes covers one change. Reading states and the TV's REST API is free.
  3. Confirm every app ID on the TV the script is for with the applications call before writing it into a script. A 404 means that ID is not installed there. IDs differ by region and model.
  4. Launch an app with a Home Assistant rest_command that sends POST to http://<tv-ip>:8001/api/v2/applications/<app-id>, called from one shared script that stops with an error when the status is not 200. Never use media_player.play_media with media_content_type: app for it, which does not open apps on a QN85D, and never the Cast entity.
  5. Turning on needs an automation on the samsungtv.turn_on trigger, because the integration sends no magic packet itself in Home Assistant 2026.10 and newer. The magic packet did not wake a QN85D from standby, a KEY_POWER did. KEY_POWER is a toggle, so the automation sends it only after GET http://<tv-ip>:8001/api/v2/ reported PowerState standby.
  6. Never print the websocket token from the config entry, and never copy it out of Home Assistant.
  7. A claim stays UNVERIFIED until it was measured on a real TV, and the mark goes with it into every note and script comment.

Related skills

  • home-assistant owns automation and script style, modes, secrets and backups
  • homelab-services owns the VM and the services that run beside Home Assistant
  • asus-router owns the router that hands the TV its address
  • The home-assistant-engineer agent owns building and changing the Home Assistant configuration
  • The security-auditor agent owns judging exposed ports and tokens

Checklist

  • Every app ID in a script answered the applications call with its name on the TV the script is for
  • Apps launch through the shared script and its rest_command POST, which stops with an error on a status other than 200, never through media_player.play_media
  • An automation on samsungtv.turn_on sends KEY_POWER only when PowerState reads standby, and the entity shows turn on
  • Power On with Mobile is on in the TV's Expert Settings, and the network type was read before trusting Wake on LAN
  • Home Assistant Cast is used only with an https external URL set
  • No address, MAC, entity ID or token of a real setup in the skill or a shared repository
  • Every change in Home Assistant had its own yes after the worst case and the undo were stated
  • Every UNVERIFIED fact used is still marked UNVERIFIED in the result

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.