Dictionary Lookup Skill

A dictionary lookup sounds like something a language model shouldn't need — until the question is "is this actually a word," "what's the IPA," or "give me a citable source," where fetched beats recalled. The Free Dictionary API answers over keyless HTTPS with definitions, phonetics, audio, and origins. This skill fetches, formats by part of speech, and keeps a clean line between what the source says and what the model adds — because a dictionary answer's whole value is knowing which is which.

What This Skill Produces

• The entry — definitions grouped by part of speech, with example sentences where the source has them
• Pronunciation — IPA text and the audio-file link when available
• Synonyms/antonyms — from the source, supplemented (and labeled) by the model when thin
• The command — exact curl, rerunnable; and an honest miss-report when the API lacks the word

Required Inputs

Ask for these if not provided:

• The word — and the sense if context suggests one ("mean" the verb, the adjective, or the statistic?)
• What they actually need — a quick meaning, pronunciation, etymology, or a citable check that a word exists — the output leads with it
• Language note — this API is English-only; other languages get an honest redirect (Wiktionary manually) rather than a fake answer

Framework: Fetch, Format, and the Honesty Line

1. The call: curl -s "https://api.dictionaryapi.dev/api/v2/entries/en/serendipity" → JSON array: phonetic/phonetics[] (IPA + .audio mp3 links), meanings[] grouped by partOfSpeech, each with definitions[] (with example sometimes), synonyms, antonyms, plus origin on some entries.
2. Lead with the asked-for thing: a pronunciation question gets IPA in line one, not after four noun senses; an is-it-a-word check gets yes/no plus the entry as proof.