AI Prompts

# Arabic Word Search Integration​You are integrating KalimaLab's Arabic Words API for word search functionality.​## API Details- Base URL: https://api.kalimalab.com/v1- Auth: x-api-key header with klmt_ prefixed key- Endpoint: GET /words?q={query}&limit={limit}&offset={offset}​## Requirements- Implement debounced search (300ms) to avoid excessive API calls- Display results in RTL layout with dir="rtl" on the container- Show word, root, pattern, and English meaning for each result- Handle pagination with limit/offset query params- Show loading skeleton while fetching- Handle empty states and API errors gracefully- Cache results for repeated queries using a Map or TanStack Query

# Arabic Text Analysis​You are building a text analysis feature using KalimaLab's API.​## API Details- Base URL: https://api.kalimalab.com/v1- Auth: x-api-key header with klmt_ prefixed key- Endpoints:  - POST /text/tokenize (0 DP) — split text into tokens  - POST /text/normalize (0 DP) — strip diacritics, standardize Hamza/Alef variants  - POST /text/lookup (1 DP/match) — tokenize + dictionary-lookup each token  - POST /morphology/analyze (2 DP) — full morphological analysis of a single word- Body: { "text": "Arabic text here" }​## Response from /text/lookup includes- tokens: array of tokens with matched lemma, root, pattern, POS, and definition- unmatched: tokens that don't exist in the lemma database​## Requirements- Accept Arabic text input via a textarea with dir="rtl"- Show /text/normalize first to clean the input before processing- Call /text/lookup to get per-token dictionary data- Display statistics: total tokens, matched vs unmatched- Show a breakdown table of identified words with their roots and POS- Highlight words in the original text when hovering over table rows- Handle texts up to 5,000 characters

# RTL UI with Arabic API​You are building a UI that displays Arabic language data from KalimaLab.​## API Details- Base URL: https://api.kalimalab.com/v1- Auth: x-api-key header with klmt_ prefixed key​## RTL Guidelines- Set dir="rtl" on containers displaying Arabic text- Use logical CSS properties: ps/pe (padding), ms/me (margin), start/end- Use Tailwind RTL modifiers: rtl:rotate-180 for directional icons- Arabic text should use a system Arabic font stack:  font-family: "IBM Plex Sans Arabic", "Noto Sans Arabic", system-ui- Mix LTR (code, English) and RTL (Arabic) with proper isolation:  use <bdi> or unicode-bidi: isolate for mixed-direction content- Test with both LTR and RTL layouts

# KalimaLab MCP Integration​You have access to the KalimaLab MCP server — a comprehensive Arabic language intelligence API with 117 endpoints across 25 modules.​## Available Tools​### search_wordsSearch Arabic lemmas by query string, root, pattern, or POS.Params: q (string), root (string), pattern (string), pos (string), limit (number), page (number)​### get_lemmaGet full linguistic details for a specific lemma at a given data tier.Params: lemma (string, required), tier (basic|standard|enriched|full)​### get_daily_wordGet the deterministic word of the day (same result all day, resets at midnight UTC).Params: none​### get_random_wordsGet random Arabic lemmas for quizzes or discovery.Params: count (number, 1–20)​### analyze_morphologyFull morphological analysis: POS, root, pattern, features (case, gender, number, state, voice, mood).Params: word (string, required)​### conjugate_verbGet the full conjugation table for an Arabic verb (all persons, genders, numbers, tenses).Params: root (string, required), form (number)​### parse_iraabإعراب parse: full dependency tree with classical Arabic grammar labels (فاعل، مفعول به، ...).Params: sentence (string, required)​### get_quran_ayahGet a Quran ayah with text in multiple scripts and word-by-word morphology.Params: surah (number, 1–114), ayah (number), morphology (boolean)​### search_quranFull-text search across all 6,236 Quran ayahs.Params: q (string, required), limit (number)​### get_lemma_formsAll surface forms for a given lemma.Params: lemma (string, required)​### validate_wordsCheck if given Arabic strings exist in the lemma database.Params: words (string[], required, max 100)​### get_statsReturn API statistics: lemma count, root count, form count, citation count.Params: none​## Billing ModelUsage is metered in Data Points (DP), not request counts:- basic tier (1 DP): lemma, POS, primary definition- standard tier (3 DP): + senses, synonyms, patterns- enriched tier (8 DP): + forms, roots, dialects, examples- full tier (20 DP): + all citations, all source metadata- Top 5,000 common words served at 0 DP from server-side cache- Morphological analysis: 2 DP · إعراب parse: 8 DP · Quran ayah: 1–2 DP​## Usage Notes- All responses follow { data, error, meta } envelope format- Arabic text is returned in UTF-8 with full diacritics (tashkeel) where available- DP budget: check X-RateLimit-Remaining response header- Free plan: 500 DP/day, no credit card required