llm-evaluation

llm-evaluation is a skill for Claude Code from ckorhonen/claude-skills. It costs 87 tokens per session (5,125 once invoked), scanned A, original, MIT.

A testing guide for AI applications that use large language models. It covers automatic scores, model-based judging, human review, and comparisons between prompts or models.

In plain words
What is it for?
Use it to evaluate retrieval-augmented generation (RAG), agent task success, generated text, classifications, and structured outputs, while tracking results over time.
Why use it?
It replaces ad hoc checking with repeatable ways to measure quality and spot regressions before release. It also helps explain unexpected model behavior.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the claude-skills plugin — 62 skills, 4 commands, 7 agents shipped together

Good fit Use it to evaluate retrieval-augmented generation (RAG), agent task success, generated text, classifications, and structured outputs, while tracking results over time.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/ckorhonen/claude-skills/llm-evaluation
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add ckorhonen/claude-skills --skill llm-evaluation
Clone the repo
git clone --depth 1 https://github.com/ckorhonen/claude-skills

Made for: Claude Code.

Or install claude-skills, the plugin that ships this one along with the rest of its 62 skills, 4 commands, 7 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for llm-evaluation

README.md
[![agentmods](https://agentmods.dev/badge/skills/ckorhonen/claude-skills/llm-evaluation/github.svg)](https://agentmods.dev/skills/ckorhonen/claude-skills/llm-evaluation)
Your own site
<a href="https://agentmods.dev/skills/ckorhonen/claude-skills/llm-evaluation"><img src="https://agentmods.dev/badge/skills/ckorhonen/claude-skills/llm-evaluation/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for llm-evaluation

Your own site · 80×15
<a href="https://agentmods.dev/skills/ckorhonen/claude-skills/llm-evaluation"><img src="https://agentmods.dev/badge/skills/ckorhonen/claude-skills/llm-evaluation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 87 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 5,125 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00087 $0.05125
Opus 5 $0.00044 $0.02563
Sonnet 5 $0.00017 $0.01025
Haiku 4.5 $0.00009 $0.00513

Measured 12d ago against content hash fa3de88c018c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

llm-evaluation scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

commit = subprocess.check_output(["git", "rev-parse", "HEAD"]).decode().strip()
skills/llm-evaluation/SKILL.md · 685 lines

How it starts

The opening of the file, as written. The whole thing — 685 lines — stays where its author put it; the contents beside it link to each section on GitHub.

LLM Evaluation

Master comprehensive evaluation strategies for LLM applications, from automated metrics to human evaluation and A/B testing.

When to Use This Skill

  • Measuring LLM application performance systematically
  • Comparing different models or prompts
  • Detecting performance regressions before deployment
  • Validating improvements from prompt changes
  • Building confidence in production systems
  • Establishing baselines and tracking progress over time
  • Debugging unexpected model behavior
  • Evaluating RAG pipeline quality (retrieval + generation)
  • Measuring agentic task success rates
  • Testing structured output schema compliance

Core Evaluation Types

1. Automated Metrics

Fast, repeatable, scalable evaluation using computed scores.

Text Generation:

  • BLEU: N-gram overlap (translation)
  • ROUGE: Recall-oriented (summarization)
  • METEOR: Semantic similarity
  • BERTScore: Embedding-based similarity
  • Perplexity: Language model confidence

Classification:

  • Accuracy: Percentage correct
  • Precision/Recall/F1: Class-specific performance
  • Confusion Matrix: Error patterns
  • AUC-ROC: Ranking quality

Retrieval (RAG):

  • MRR: Mean Reciprocal Rank
  • NDCG: Normalized Discounted Cumulative Gain
  • Precision@K: Relevant in top K
  • Recall@K: Coverage in top K

2. Human Evaluation

Manual assessment for quality aspects difficult to automate.

Dimensions:

  • Accuracy: Factual correctness
  • Coherence: Logical flow
  • Relevance: Answers the question
  • Fluency: Natural language quality
  • Safety: No harmful content
  • Helpfulness: Useful to the user

3. LLM-as-Judge

Use stronger LLMs to evaluate weaker model outputs. This is the dominant approach in 2025/2026 for open-ended tasks.

Approaches:

  • Pointwise: Score individual responses (0-10 Likert scales)
  • Pairwise: Compare two responses (preferred by MT-Bench, Chatbot Arena)
  • Reference-based: Compare to gold standard answer
  • Reference-free: Judge without ground truth (good for creative/open-ended tasks)
  • Rubric-based: Judge against explicit criteria (best for consistency)
  • Constitutional: Check against a set of principles or rules

Read the full file on GitHub · 685 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 685 lines · 87 tokens per session scan A fa3de88c018c

Subscribe to this mod's changes

llm-evaluation is a skill published in the GitHub repository ckorhonen/claude-skills (14 stars, last pushed 2mo ago), licensed MIT. It adds 87 tokens to every session and 5,125 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

llm-as-judge-evaluation

Evaluate LLM outputs using frontier models as judges. Use for pairwise model comparison, quality scoring with custom rubrics, and automated evaluation pipelines. Covers position bias mitigation, statistical significance, and generating preference data for DPO/RLHF.

synthetic-sciences/openscience · 56 tokens

autoresearch

Autonomous goal-directed iteration loop that continuously improves prompts, templates, configs, or code. Two evaluation modes — deterministic (eval.py with proxy heuristics) or AI judge (LLM rubric scoring). Uses four-way separation in both modes. Inspired by Karpathy's autoresearch.

naveedharri/benai-skills · 59 tokens

cli-eval

Create and run evaluation suites, watch live benchmark progress, view scorecards, compare model performance, and integrate eval runs with CI workflows from the CLI.

diegosouzapw/OmniRoute · 34 tokens

model-merging

Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task…

davila7/claude-code-templates · 73 tokens

darwinian-evolver

Evolve prompts/regex/SQL/code with Imbue's evolution loop.

NousResearch/hermes-agent · 22 tokens

validate

Validate Semantica pipelines, extraction quality, graph schemas, and ontology consistency. Returns structured error/warning checklists. Uses PipelineValidator, PipelineBuilder.validatepipeline(), GraphValidator, and OntologyValidator. Sub-commands: pipeline, step, dependencies, extraction, graph, ontology, performance.

semantica-agi/semantica · 0 tokens