golden-dataset

golden-dataset is a skill for Claude Code from ArieGoldkin/claude-forge. It costs 25 tokens per session (1,262 once invoked), scanned A, original, MIT.

A guide to maintaining a golden dataset: a carefully checked collection of examples used to test and compare AI or search systems. It covers adding entries, backups, validation, duplicate detection, and coverage checks.

In plain words
What is it for?
Use it to prepare evaluation examples, validate data schemas, restore backups, find duplicates, measure coverage, test search, and compare embedding models.
Why use it?
It gives you a stable reference for finding regressions and checking whether changes affect retrieval quality or model results.

Skill for Claude Code

Written for Claude Code: paths in frontmatter.

Part of the atk plugin — 16 skills, 25 commands, 1 agent, 1 hook shipped together

Good fit Use it to prepare evaluation examples, validate data schemas, restore backups, find duplicates, measure coverage, test search, and compare embedding models.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/ariegoldkin/claude-forge/golden-dataset
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 ArieGoldkin/claude-forge --skill golden-dataset
Clone the repo
git clone --depth 1 https://github.com/ArieGoldkin/claude-forge

Made for: Claude Code.

Or install atk, the plugin that ships this one along with the rest of its 16 skills, 25 commands, 1 agent, 1 hook.

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 golden-dataset

README.md
[![agentmods](https://agentmods.dev/badge/skills/ariegoldkin/claude-forge/golden-dataset/github.svg)](https://agentmods.dev/skills/ariegoldkin/claude-forge/golden-dataset)
Your own site
<a href="https://agentmods.dev/skills/ariegoldkin/claude-forge/golden-dataset"><img src="https://agentmods.dev/badge/skills/ariegoldkin/claude-forge/golden-dataset/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 golden-dataset

Your own site · 80×15
<a href="https://agentmods.dev/skills/ariegoldkin/claude-forge/golden-dataset"><img src="https://agentmods.dev/badge/skills/ariegoldkin/claude-forge/golden-dataset.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 25 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,262 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00025 $0.01262
Opus 5 $0.00013 $0.00631
Sonnet 5 $0.00005 $0.00252
Haiku 4.5 $0.00003 $0.00126

Measured 10d ago against content hash 0f23bdb365c6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

golden-dataset scanned grade A with 0 findings 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 10d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/ai-toolkit/skills/golden-dataset/SKILL.md · 123 lines

How it starts

The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Golden Dataset

Curate, manage, and validate high-quality test datasets for AI/ML systems

Overview

A golden dataset is a curated collection of high-quality examples used for regression testing, retrieval evaluation, model benchmarking, and reproducibility. This skill covers the full lifecycle: curating new entries with quality analysis, managing backup/restore operations, and validating data integrity.

Example Golden Dataset Metrics

Metric Value
Documents 98 completed
Chunks 415 embedded segments
Test queries 203 with expected results
Pass rate 91.6% retrieval quality

Purpose: Test hybrid search (vector + BM25 + RRF), validate metadata boosting, detect retrieval regressions, benchmark embedding models.


Curation

Quality criteria, workflows, and multi-agent analysis patterns for evaluating and adding documents to the golden dataset.

Key areas:

  • Content type classification -- article, tutorial, research paper, documentation, video transcript, code repository
  • Difficulty stratification -- trivial, easy, medium, hard, adversarial (based on semantic complexity)
  • Quality dimensions -- accuracy (0.25), coherence (0.20), depth (0.25), relevance (0.30)
  • Multi-agent pipeline -- parallel evaluation with Quality Evaluator, Difficulty Classifier, Domain Tagger, Query Generator
  • Duplicate prevention -- URL check + semantic similarity > 80% threshold

Detailed patterns: references/curation.md Multi-agent pipeline: references/multi-agent-pipeline.md


Management

Backup, restore, and lifecycle operations for protecting golden dataset integrity.

Key areas:

  • Data integrity contracts -- real canonical URLs required, no placeholders
  • JSON backup strategy -- version-controlled, human-readable, portable; embeddings excluded and regenerated on restore
  • Restore process -- load JSON, validate structure, create analyses/chunks, regenerate embeddings, verify integrity
  • CLI usage -- poetry run python scripts/backup_golden_dataset.py backup|verify|restore
  • CI/CD automation -- GitLab CI pipeline integration for scheduled backups

Read the full file on GitHub · 123 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. 10d ago First seen · 123 lines · 25 tokens per session scan A 0f23bdb365c6

Subscribe to this mod's changes

golden-dataset is a skill published in the GitHub repository ArieGoldkin/claude-forge (6 stars, last pushed 1mo ago), licensed MIT. It adds 25 tokens to every session and 1,262 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

spark-environment-setup

Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.

wshobson/agents · 76 tokens

spark-memory-thermal-ops

Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.

wshobson/agents · 59 tokens

spark-training-gotchas

Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.

wshobson/agents · 63 tokens

langchain

Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, Anthropic, Google), 500+ integrations, ReAct agents, tool calling, memory management, and vector store retrieval. Use for building chatbots, question-answering systems, autonomous agents, or RAG…

davila7/claude-code-templates · 79 tokens

llamaindex

Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing, and querying. Features vector indices, query engines, agents, and multi-modal support. Use for document Q&A, chatbots, knowledge retrieval, or building RAG pipelines. Best for data-centric LLM…

davila7/claude-code-templates · 70 tokens

sglang

Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic workflows with tool calls, or when you need 5× faster inference than vLLM with prefix sharing. Powers 300,000+ GPUs at xAI, AMD, NVIDIA, and LinkedIn.

davila7/claude-code-templates · 72 tokens