ai-data-curation-pretraining

ai-data-curation-pretraining is a skill for Codex from vasilyu1983/AI-Agents-public. It costs 48 tokens per session (4,614 once invoked), scanned A, original, MIT.

A workflow for preparing large collections of text for training language models. It covers collecting text, extracting it from web pages, identifying languages, filtering low-quality or repeated content, and checking for overlap with evaluation tests.

In plain words
What is it for?
Use it to build or review pretraining datasets, choose filtering and deduplication methods, remove evaluation-set contamination, and run controlled comparisons of data choices.
Why use it?
Raw web and synthetic text can contain duplicates, noise, unwanted languages, or examples copied from test sets. The workflow provides specific processing stages for cleaning and auditing the training data.

Skill for Codex

Written for Codex: agents/openai.yaml present. Also seen: mentions Claude Code; mentions Codex.

Good fit Use it to build or review pretraining datasets, choose filtering and deduplication methods, remove evaluation-set contamination, and run controlled comparisons of data choices.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining
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 vasilyu1983/AI-Agents-public --skill ai-data-curation-pretraining
Clone the repo
git clone --depth 1 https://github.com/vasilyu1983/AI-Agents-public

Made for: Codex.

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 ai-data-curation-pretraining

README.md
[![agentmods](https://agentmods.dev/badge/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining/github.svg)](https://agentmods.dev/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining)
Your own site
<a href="https://agentmods.dev/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining"><img src="https://agentmods.dev/badge/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining/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 ai-data-curation-pretraining

Your own site · 80×15
<a href="https://agentmods.dev/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining"><img src="https://agentmods.dev/badge/skills/vasilyu1983/ai-agents-public/ai-data-curation-pretraining.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 48 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,614 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. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 $0.00048 $0.04614
Opus 5 $0.00024 $0.02307
Sonnet 5 $0.00010 $0.00923
Haiku 4.5 $0.00005 $0.00461

Measured 8d ago against content hash 24c0e7480fe6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

ai-data-curation-pretraining 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 8d 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.

frameworks/shared-skills/skills/ai-data-curation-pretraining/SKILL.md · 255 lines

How it starts

The opening of the file, as written. The whole thing — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Pretraining Data Curation — Functional Reference Skill

Domain: Building web-scale and synthetic pretraining corpora, running controlled data ablations. Distinct from applications-layer retrieval (RAG) and general data engineering.

No theory. No generic pipeline intros. Focus on stage-by-stage decisions, heuristic thresholds, tooling choices, and ablation protocol.

ASCII Flow

CommonCrawl WARCs
  |
  v
[Extract]  trafilatura / datatrove HTMLExtractor
  raw text + metadata (URL, timestamp, content-type)
  |
  v
[Language ID]  fastText lid.176.bin
  keep target language(s), threshold ≥ 0.65
  |
  v
[Quality Filter — Heuristic]  Gopher / C4 rules
  symbol-to-word ratio, fraction lines ending ellipsis,
  stopword density, word count bounds, mean word length
  |
  v
[Quality Filter — Classifier]  FineWeb-Edu edu-score / custom
  trained on human labels; outperforms heuristics on recall
  |
  v
[Near-Dedup]  MinHash + LSH banding (datasketch)
  n-gram shingles -> MinHash signature -> band partitioning
  |
  v
[Exact-Substring Dedup]  suffix-array substring match
  remove exact repeated sequences across documents
  |
  v
[Decontamination]  n-gram match against eval benchmarks
  FAIL LOUD — contaminated eval numbers are the field's #1 silent failure
  |
  v
[PII / Safety Scrub]  regex + classifier
  email, phone, SSN, credit card patterns; hate/CSAM removal
  |
  v
[Tokenize + Shard]  HF tokenizers / tiktoken; Parquet shards
  |
  v
[Domain Mix + Weight]  dolma toolkit / custom sampling
  web / books / code / math / synthetic — proportions are a research lever
  |
  v
[Train + Eval]  nanotron / lighteval / lm-evaluation-harness
  ablation output: eval delta per pipeline stage

When to Use This Skill

Activate when the task involves:

  • Finding existing high-quality datasets for pretraining or fine-tuning before building from scratch (see Dataset Discovery reference)
  • Sourcing and filtering CommonCrawl WARCs or other web-scale corpora
  • Implementing or debugging any stage of the curation pipeline above
  • Designing quality filters (heuristic or classifier-based)
  • Running MinHash / LSH deduplication or exact-substring dedup
  • Decontaminating a dataset against evaluation benchmarks
  • Generating synthetic pretraining data (Cosmopedia, Self-Instruct, Evol-Instruct, Nemotron)
  • Designing and executing controlled data ablations
  • Writing datasheets (Gebru et al.) for a curated dataset
  • Understanding open recipe datasets: FineWeb, Dolma, The Pile, RedPajama, SlimPajama, C4, RefinedWeb, OLMo

Read the full file on GitHub · 255 lines

Files

What ships with it

7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 8d ago Changed · +2 lines 24c0e7480fe6
  2. 12d ago First seen · 253 lines · 48 tokens per session scan A dc898c11390c

Subscribe to this mod's changes

ai-data-curation-pretraining is a skill published in the GitHub repository vasilyu1983/AI-Agents-public (87 stars, last pushed 9d ago), licensed MIT. It adds 48 tokens to every session and 4,614 once invoked, about $0.0002 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 2026-08-30.

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

data-quality-frameworks

Implement data quality validation with Great Expectations, dbt tests, and data contracts. Use when building data quality pipelines, implementing validation rules, or establishing data contracts.

wshobson/agents · 37 tokens

airflow-dag-patterns

Build production Apache Airflow DAGs with best practices for operators, sensors, testing, and deployment. Use when creating data pipelines, orchestrating workflows, or scheduling batch jobs.

wshobson/agents · 42 tokens

dbt-transformation-patterns

Master dbt (data build tool) for analytics engineering with model organization, testing, documentation, and incremental strategies. Use when building data transformations, creating data models, or implementing analytics engineering best practices.

wshobson/agents · 47 tokens