Lean and efficient with no padding or explanation of concepts Claude already knows; every section (Trigger, Action, Procedure, Example, Notes) earns its place.

Provides the exact executable output format 'Action: open <receptacle_name>' plus a complete worked example with observation, thought, and action — copy-paste ready.

The single action is unambiguous and sequenced as a clear two-step procedure; the simple-skill exception applies and no validation checkpoint is required for a non-destructive single open action.

Under 50 lines with no need for external references, organized into clear labeled sections; meets the bar for a well-organized self-contained skill.

Names the domain and concrete actions ('outputs the open action', 'cooling an item with a fridge', 'taking an object from inside a cabinet'), but it is essentially a single action with use-cases rather than multiple distinct concrete actions.

Clearly answers both 'what' ('opens a closed receptacle', 'outputs the open action') and 'when' with an explicit trigger clause ('It should be triggered when the agent encounters a closed receptacle...that must be opened to proceed with a task').

Quotes the exact natural trigger pattern ('The fridge 1 is closed.') and lists common receptacle terms (fridge, cabinet, safe), giving good coverage of terms the environment would actually produce.

The 'The X is closed' trigger carves out a clear niche for opening receptacles, and adjacent actions (take, cool, put) are explicitly framed as separate skills, making conflicts unlikely.