# Vnthuquan

> Use when the user explicitly wants Vietnam Thu Quan / vnthuquan / vietnamthuquan.eu ebook discovery, metadata, mirror checks, categories, formats, archive inspection, or controlled downloads through the local vnthuquan package.

- Skill: `hoanganhduc/vnthuquan` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add hoanganhduc/vnthuquan`
- Raw SKILL.md: https://api.skillmd.com/api/skills/hoanganhduc/vnthuquan/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: hoanganhduc (https://skillmd.com/u/hoanganhduc)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/hoanganhduc/vnthuquan

---


# vnthuquan


## Windows Runtime Commands

On native Windows, use the managed Windows runner and the native runtime command target. Set `$runtime` to the installed runtime root. Multi-agent installs usually use `%LOCALAPPDATA%\ai-agents-skills\runtime`. Then run:

```powershell
$runtime = if ($env:AAS_RUNTIME_ROOT) { $env:AAS_RUNTIME_ROOT } else { "$env:LOCALAPPDATA\ai-agents-skills\runtime" }
& "$runtime\run_skill.ps1" "skills/vnthuquan/run_vnthuquan.ps1" <args>
& "$runtime\run_skill.ps1" "skills/vnthuquan/run_vnthuquan.ps1" <args>
```

POSIX examples below use `run_skill.sh` and `.sh` command targets; use the Windows command target above on native Windows.

Use this skill only for Vietnam Thu Quan site-specific ebook work:

- search Vietnam Thu Quan books
- list categories, formats, latest books, authors, ranked lists, or category
  contents
- inspect book metadata and shareable source links
- check mirrors and wrapper/package health
- inspect the wrapper-managed download archive
- controlled dry-run/executed downloads, queue execution, validation, failure
  recovery, and guarded Calibre handoff

## Routing Boundary

- Do not use this skill as the first route for generic papers, DOI, ISBN, or
  book retrieval. Follow the library-first routing rules: Zotero first, then
  Calibre or getscipapers as appropriate.
- Do not replace the `calibre` skill for local library management. Use Calibre
  only after a `vnthuquan` download has been validated or when the user
  explicitly asks for Calibre integration.

## Runtime

Use the managed runtime runner rather than invoking the wrapper directly:

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh diagnose --json
```

Runtime wrapper:

- `$AAS_RUNTIME_WORKSPACE/skills/vnthuquan/run_vnthuquan.sh`

Detailed workflows live in:

- `references/workflows.md`

## Current Phase

Phase 5 is cross-target discovery, controlled downloads, queue management,
validation, Calibre dry-run handoff, and guarded Calibre write integration. It
includes:

- `diagnose`
- `doctor`
- `mirrors list`
- `mirrors check`
- `config path`
- `config show`
- `categories list`
- `categories show`
- `formats`
- `list ...`
- `search ...`
- `show ...`
- `archive path`
- `archive list`
- `completion ...`
- download dry-run by default
- download execution only with `--execute --yes`
- queue manifest creation
- queue execution only with `execute-queue ... --yes`
- failed-queue retry manifest creation
- local file validation
- Calibre dry-run handoff for validated EPUB/PDF files
- Calibre writes only through
  `add-to-calibre PATH --execute --yes --duplicates-reviewed`

## Examples

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh search "Kim Dung" --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh categories list --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh show --title "Mưa Đỏ" --links --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh mirrors check --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh download --title "Mưa Đỏ" --format epub --dry-run --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh queue --query "Kim Dung" --limit 5 --format epub --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh add-to-calibre PATH --dry-run --json
```

```bash
bash "${AAS_RUNTIME_ROOT:-$HOME/.local/share/ai-agents-skills/runtime}/run_skill.sh" skills/vnthuquan/run_vnthuquan.sh add-to-calibre PATH --execute --yes --duplicates-reviewed --json
```

## Safety Rules

- Prefer `--json` for machine-readable output.
- If multiple books match, show numbered candidates and ask the user to choose.
- Keep wrapper config, cache, and archive state under the target-local Codex
  state directory.
- Do not mutate package default config unless the user explicitly asks for that
  after seeing the target path.
- Do not execute downloads or queue manifests unless the command includes the
  wrapper confirmation flag required for that operation.
- Do not perform Calibre writes unless a dry-run has been reviewed, duplicate
  candidates have been checked, and the command includes `--execute --yes
  --duplicates-reviewed`. Add `--allow-duplicate` only when duplicate
  candidates are intentionally being accepted as a separate Calibre entry.