Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Major update: added vowel symbols reference table, shva rules, complete begedkefet table (6 letters with modern Hebrew notes), all 7 binyanim (added Pu'al and Huf'al), preposition nikud rules, more homographs and foreign names, common mistakes section.
What This Skill Does
Hebrew nikud rules for AI agents adding vowel points to Hebrew text before TTS synthesis. Covers the seven verb binyanim, dagesh in begedkefet letters, gender suffixes, homographs, and foreign loanwords. Includes a decision tree for when to add nikud and when to omit it.
Provides explicit rules against over-nikuding, preventing TTS from reading wrong vowels literally instead of inferring correct pronunciation from context.
When to use it
- Adding pronunciation hints to Hebrew TTS scripts
- Disambiguating homographs like ספר (book, counted, or told)
- Marking correct dagesh in transliterated foreign names
- Applying gender-correct pronoun suffixes in Hebrew dialogue
- Choosing the right binyan nikud pattern for a Hebrew verb
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: add nikud to 'בקר טוב, סיפרתי לה על הפיצה בפריז'
- 1Skip nikud on common unambiguous words like טוב
- 2Identify בקר as homograph, select בּוֹקֶר (morning) based on context
- 3Look up סיפרתי in the Pi'el conjugation table and mark as סִפַּרְתִּי
- 4Apply dagesh to פ in פיצה and פריז as foreign loanwords requiring P not F sound
בּוֹקֶר טוב, סִפַּרְתִּי לה על הפִּיצָה בְּפָרִיז