Evaluate Arabic tool-calling performance of LLMs with structured benchmarks.
npx skills add https://github.com/Moshe-ship/mkhlab --skill arabic-agent-eval
Please help me install this Agent Skill. Skill: arabic-agent-eval Source: https://github.com/Moshe-ship/mkhlab/tree/main/hermes-skills/arabic-agent-eval Command: npx skills add https://github.com/Moshe-ship/mkhlab --skill arabic-agent-eval
Evaluates how effectively LLMs handle Arabic tool calling, providing a repeatable benchmark to measure capability and reliability.
Run a full Arabic-agent-eval benchmark to compare tool calling across models.
High-intent search queries and answers about installing and using this skill.
To benchmark Arabic tool-calling performance, run a structured evaluation against a predefined dataset to measure LLM capability and reliability. This skill enforces prerequisites and outputs structured metrics for function calling and parameter extraction.
Arabic tool-calling evaluation measures how effectively LLMs handle function invocation and parameter extraction in Arabic. It provides a repeatable benchmark across six evaluation categories to identify gaps in prompt handling.
Yes, this Arabic tool-calling benchmark supports evaluating multiple Arabic dialects. It tests LLMs on tasks including function calling and parameter extraction across dialects to provide cross-model comparisons and error analysis.
The best way to compare Arabic prompt handling is running a full benchmark workflow that generates structured reporting. It provides cross-model comparisons and qualitative insights to identify reliability gaps in Arabic tool calling.
You can start an Arabic LLM evaluation benchmark using the quick-start workflow to run a full evaluation. The skill enforces prerequisites before running structured tasks across six evaluation categories.
The Arabic LLM benchmark covers six evaluation categories, supporting multiple Arabic dialects. It tests function calling and parameter extraction, outputting structured metrics and qualitative insights for cross-model error analysis.
Moshe-shipCollect and clean Arabic web content into JSONL training data.
paimlAnalyze polyglot codebases to detect languages, assess distribution, and generate per-language reports.
robotics-playgroundRefine LLM output through deterministic validation and refinement steps.
jackquelinunpredictable827Benchmark Arabic tool-calling performance for language models using the aae CLI.
