# Repository Resource Portal

> Maintain this repository as a growing educational LLM knowledge base. Use when Codex needs to add, reorganize, or review homepage navigation, localized landing pages, directory-level catalogs, dataset and guide links, Codex Goal templates, or resource-portal documentation while preserving concise root navigation and relative links. Use when this capability is needed.

- Skill: `tomevault-io/repository-resource-portal` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/repository-resource-portal`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/repository-resource-portal/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/repository-resource-portal

---


# Repository Resource Portal

Use this skill when changing repository navigation or knowledge-base catalogs.

## Maintenance Rules

- Keep the root README concise and navigational.
- Preserve four-language landing-page backups before major rewrites.
- Add new content to directory-level catalogs first.
- Use moderate, consistent emojis.
- Preserve relative links.
- Update English, Chinese, Korean, and Japanese landing pages when homepage navigation changes.
- Keep detailed tool documentation inside the relevant subproject.
- Avoid duplicating the full MTP pipeline explanation in the root README.
- Add new Codex Goal workflows and Skills in both `.agents/skills/` and `codex-goals/` when they are meant to be reusable.
- Run repository-relative link checks and privacy checks before any sync.

## Canonical Locations

- Root navigation: `README.md`
- Localized landing pages: `docs/README_zh.md`, `docs/README_ko.md`, `docs/README_ja.md`
- Goal catalog: `codex-goals/`
- Repository Skills: `.agents/skills/`
- MTP implementation: `qwen-mtp-gguf/`
- Training notebooks: `train_code/`
- Dataset folders: `High-fidelity Dataset/`
- PDF guides: `guidePDF/`

---
> Source: [R6410418/Jackrong-llm-finetuning-guide](https://github.com/R6410418/Jackrong-llm-finetuning-guide) — distributed by [TomeVault](https://tomevault.io).
<!-- tomevault:4.0:skill_md:2026-06-18 -->

