Skip to main content
Skills Hub

Best LLM Evaluation Skills 2026

Curated skills for evaluating LLM-powered applications: OpenAI Evals, Promptfoo, Ragas, DeepEval, LangSmith, LangChain Evaluators, Helicone, Arize Phoenix, TruLens, Evidently AI, Weights & Biases. Install with a single command for Claude Code, Cursor, Copilot, and 30+ agents.

Install in 5 seconds:

# Install LLM eval skills
npx @qaskills/cli add promptfoo-llm-evals
npx @qaskills/cli add openai-evals
npx @qaskills/cli add ragas-rag-evals

Multi-provider

Promptfoo, OpenAI Evals, Ragas, DeepEval — covers OpenAI, Anthropic, local models.

RAG-ready

Context precision, recall, faithfulness, answer relevance metrics built in.

Red-teaming

Promptfoo red-team skill covers jailbreaks, prompt injection, PII leakage.

Top 7 Skills

Ranked by install count. All quality-scored 0-100.

RAG Regression Testing

by thetestingacademy

80

Gate RAG pipelines in CI with versioned golden eval sets, per-metric thresholds, baseline drift detection, and a build that fails when retrieval or answer quality regresses.

regression
llm-evals
integration
ragas
deepeval
1780/100

Promptfoo LLM Red Teaming

by thetestingacademy

88

Evaluate and red-team LLM applications with promptfoo, declarative YAML evals, assertions, model comparisons, and automated adversarial scans for prompt injection, jailbreaks, PII leaks, and unsafe outputs in CI.

llm-evals
security
regression
promptfoo
1688/100

Langfuse LLM Observability Testing

by thetestingacademy

80

Instrument LLM apps with Langfuse tracing, then use traces, scores, and datasets to test in production, run evaluations on real traffic, catch regressions, and close the loop from incident to golden dataset.

llm-evals
integration
regression
langfuse
1680/100

Ragas RAG Evaluation

by thetestingacademy

80

Evaluate RAG pipelines with Ragas, measuring faithfulness, answer relevancy, context precision and recall, building golden datasets, and wiring threshold gates into CI for retrieval regressions.

llm-evals
integration
regression
ragas
pytest
1580/100

OpenAI Evals Trace Grading

by thetestingacademy

88

Grade LLM and agent traces with OpenAI Evals - build datasets, configure string/python/model graders, run eval suites, and gate agent behavior changes in CI.

llm-evals
regression
integration
openai-evals
pytest
1388/100

DeepEval LLM Evaluation

by thetestingacademy

88

Test LLM applications with DeepEval, pytest-style unit tests for LLM outputs using G-Eval, answer relevancy, faithfulness, hallucination and custom metrics, with CI quality gates and dataset-driven regression runs.

llm-evals
unit
regression
deepeval
pytest
588/100

RAG Evaluation Metrics

by thetestingacademy

82

Measure RAG pipeline quality with context precision/recall, faithfulness, answer relevancy, and groundedness using Ragas and DeepEval, with golden datasets and pass/fail thresholds.

llm-evals
integration
regression
ragas
deepeval
582/100

Frequently Asked Questions

What is LLM evaluation?

Measuring how well an LLM-powered application performs across correctness, faithfulness, factuality, safety, latency, cost. Critical for production LLM apps.

OpenAI Evals or Promptfoo?

Promptfoo for multi-provider + CI + red-teaming. OpenAI Evals for OpenAI-only Python research workflows.

What is Ragas?

RAG-focused eval library. Measures context precision, context recall, faithfulness, answer relevance for retrieval-augmented generation pipelines.

Red teaming?

Adversarial testing — jailbreaks, prompt injection, PII leakage, hallucination probes. Promptfoo has first-class red-team support.

How do skills install?

Run `npx @qaskills/cli add <skill-name>`. CLI detects your agent and writes the skill to the correct path.

Free?

Yes — MIT licensed skills. The underlying tools (Promptfoo, Ragas, DeepEval) are also OSS.

Pytest integration?

DeepEval is pytest-style natively. Ragas and Promptfoo can run inside pytest with a custom fixture.

Ready to ship better tests?

Install your first skill in 5 seconds. Browse all 500+ skills or jump straight into the recommended starter.