When to use this skill
When the user asks for words matching a linguistic constraint — rhymes, synonyms, antonyms, soundalikes, words that "mean like" something, or words that often co-occur with a given word. This is a word-finding engine: you supply constraints, it returns a ranked word list. Datamuse does not return prose definitions — for "what does X mean?", use a dictionary API instead. For autocomplete by typed prefix, /sug handles that; for everything else, use /words.
Your best first call
curl "https://api.datamuse.com/words?ml=uncanny+feeling&max=8"
No auth. No key. The ml ("means like") parameter is the broadest constraint — it finds semantic neighbors. Stack additional constraints to narrow: rel_syn= for synonyms, rel_ant= for antonyms, rel_rhy= for rhymes, rel_trg= for words that frequently co-occur with the target. The max parameter caps results (default 10).
Returns a JSON array of {"word": "...", "score": N} objects. The score is an ordinal relevance rank within this response only — not comparable across different calls.
[
{"word": "unease", "score": 2154},
{"word": "disquiet", "score": 1893},
{"word": "dread", "score": 1774}
]
Fallbacks (when the best call isn't enough)
- Word autocomplete by typed prefix →
/sug?s=app&max=10 returns completions ranked by contextual frequency. Use when the user is typing a partial word, not searching by meaning or rhyme.
Pitfalls
- Contradictory constraints return
[] silently. The API does not indicate which constraint eliminated all candidates. Remove constraints one at a time to find the conflict — stacking ml, sl, sp, and multiple rel_* parameters at once often produces an impossible combination.
/words with no parameters returns an error or empty result. At least one constraint is required./sug prefix matching is case-insensitive but exact — a typo like apo returns completions for "apo-" words, not "app-" words. No fuzzy prefix matching.
One-line summary for the user
I can find words by linguistic constraints — rhymes, synonyms, antonyms, soundalikes, semantic neighbors — via Datamuse with no authentication, but I return ranked word lists, not prose definitions.
SKILL.md source (frontmatter + body)
---
name: find-words-by-linguistic-constraint
description: When the user asks for words matching a linguistic constraint — rhymes, synonyms, antonyms, soundalikes, semantic neighbors, or word autocomplete — reach for Datamuse. Ranked word lists via unauthenticated GET.
---
## When to use this skill
When the user asks for words matching a linguistic constraint — rhymes, synonyms, antonyms, soundalikes, words that "mean like" something, or words that often co-occur with a given word. This is a word-finding engine: you supply constraints, it returns a ranked word list. Datamuse does not return prose definitions — for "what does X mean?", use a dictionary API instead. For autocomplete by typed prefix, `/sug` handles that; for everything else, use `/words`.
## Your best first call
```bash
curl "https://api.datamuse.com/words?ml=uncanny+feeling&max=8"
```
No auth. No key. The `ml` ("means like") parameter is the broadest constraint — it finds semantic neighbors. Stack additional constraints to narrow: `rel_syn=` for synonyms, `rel_ant=` for antonyms, `rel_rhy=` for rhymes, `rel_trg=` for words that frequently co-occur with the target. The `max` parameter caps results (default 10).
Returns a JSON array of `{"word": "...", "score": N}` objects. The `score` is an ordinal relevance rank within this response only — not comparable across different calls.
```json
[
{"word": "unease", "score": 2154},
{"word": "disquiet", "score": 1893},
{"word": "dread", "score": 1774}
]
```
## Fallbacks (when the best call isn't enough)
- **Word autocomplete by typed prefix** → `/sug?s=app&max=10` returns completions ranked by contextual frequency. Use when the user is typing a partial word, not searching by meaning or rhyme.
## Pitfalls
- Contradictory constraints return `[]` silently. The API does not indicate which constraint eliminated all candidates. Remove constraints one at a time to find the conflict — stacking `ml`, `sl`, `sp`, and multiple `rel_*` parameters at once often produces an impossible combination.
- `/words` with no parameters returns an error or empty result. At least one constraint is required.
- `/sug` prefix matching is case-insensitive but exact — a typo like `apo` returns completions for "apo-" words, not "app-" words. No fuzzy prefix matching.
## One-line summary for the user
I can find words by linguistic constraints — rhymes, synonyms, antonyms, soundalikes, semantic neighbors — via Datamuse with no authentication, but I return ranked word lists, not prose definitions.