Winnershitram

LLM Eval — AI skill for Claude Code

AI community

model: claude-sonnet-4-6 domain: datascience tags: ["data-science", "machine-learning", "analytics"] source_repo: BehiSecc/awesome-claude-skills 🤖 Llm Eval LLM output evaluation harness: hallucination rate….

How to install LLM Eval

Installs to ~/.claude/skills/winnershitram-r17-behisecc-awesome-claude-skills-datascience-llm-eval/SKILL.md

Terminal

mkdir -p ~/.claude/skills/winnershitram-r17-behisecc-awesome-claude-skills-datascience-llm-eval && curl -fsSL https://raw.githubusercontent.com/Winnershitram/r17-behisecc-awesome-claude-skills-datascience/HEAD/commands/llm-eval.md -o ~/.claude/skills/winnershitram-r17-behisecc-awesome-claude-skills-datascience-llm-eval/SKILL.md

Restart Claude Code, or start a new session, for it to be picked up.

What LLM Eval does

model: claude-sonnet-4-6 domain: datascience tags: ["data-science", "machine-learning", "analytics"] source_repo: BehiSecc/awesome-claude-skills

🤖 Llm Eval

LLM output evaluation harness: hallucination rate, faithfulness and latency

Context

You are a senior Data Science & AI/ML specialist. The user needs help with **llm eval** in the context of data pipelines, model training, evaluation, mlops and analytical reporting.

Requirements

$ARGUMENTS

Instructions

Step 1

Alternatives in AI

WeKnora — Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, an 26.9k ★
Opencodex — Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama… 12.3k ★
Firecrawl MCP Server — 🔥 Official Firecrawl MCP Server - Adds powerful web scraping and search to Cursor, Claude and any other LLM c 6.1k ★