MigoXLabMigoXLabOfficialยท2 Agent Skills Included

dingo

Evaluate AI training data quality and fact-check articles automatically

Evaluates training datasets, fine-tuning data, and RAG systems with 100+ rule-based and LLM-based quality metrics. Detects hallucinations, PII leaks, formatting errors, and factual errors without manual data review. Fact-checks articles claim by claim using web and academic search, then generates clear quality reports.
npx skills add MigoXLab/dingo --all -g -y
Available:

Gives the AI agent a full map of Dingo's architecture, evaluator registration patterns, data formats, and execution modes so it can run and extend data quality evaluations correctly.

All Skills in This Repository (2)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install Dingo?โ–ผ

Run `npx skills add MigoXLab/dingo --all -g -y` in your terminal to install all skills in this suite globally.

How to check training data quality with AI?โ–ผ

Dingo applies 100+ built-in rule and LLM evaluators to your JSONL, CSV, or Parquet datasets and reports issues like duplicates, PII, and low-quality text.

Can AI fact-check an article automatically?โ–ผ

Yes. The dingo-verify skill extracts every factual claim from an article and verifies each one against web and ArXiv search results, returning an accuracy score.

Does Dingo work with Claude Code and Cursor?โ–ผ

Yes. Dingo ships a built-in MCP server and standard SKILL.md files, so it runs directly inside Claude Code, Cursor, and other AI agents.

Do I need an API key to use Dingo?โ–ผ

No for rule-based checks, which run locally for free. An OpenAI-compatible API key is only needed for LLM-based evaluation and fact-checking.

Related Repositories in Data & Analytics

View All in Data & Analyticsโ†’