stardew-wiki-advisor

Query Stardew Valley Wiki using natural language. Ask about crops, NPCs, strategies, and more.

16 stars

Best use case

stardew-wiki-advisor is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Query Stardew Valley Wiki using natural language. Ask about crops, NPCs, strategies, and more.

Teams using stardew-wiki-advisor should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/stardew-wiki-advisor/SKILL.md --create-dirs "https://raw.githubusercontent.com/diegosouzapw/awesome-omni-skill/main/skills/data-ai/stardew-wiki-advisor/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/stardew-wiki-advisor/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How stardew-wiki-advisor Compares

Feature / Agentstardew-wiki-advisorStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Query Stardew Valley Wiki using natural language. Ask about crops, NPCs, strategies, and more.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Stardew Valley Wiki Advisor

問牧場物語的任何問題。由本機 AI 和本地 Wiki 向量資料庫驅動。

## Setup (First Time Only)

在使用前,需要初始化向量資料庫。

```bash
# 1. 安裝依賴
pip install requests beautifulsoup4 lxml numpy faiss-cpu ollama

# 2. 下載 Embedding 模型
ollama pull qwen3-embedding

# 3. 下載 LLM 模型
ollama pull qwen3:8b

# 4. 爬取 Wiki (100 頁)
python3 {baseDir}/scripts/crawl_wiki.py --max-pages 100 --output {baseDir}/data

# 5. 建立向量索引
python3 {baseDir}/scripts/build_vectors.py --input {baseDir}/data/raw_pages.json --output {baseDir}/data --model qwen3-embedding

# 初始化完成!
```

Takes ~20-30 minutes on first run:
- 5 min: crawl 100 Wiki pages
- 12-15 min: vectorization (100 pages × 10 chunks = 1000+ embeddings)
- Ollama model downloads: ~10-20 min (first time only)

## Query

> Note (OpenClaw sandbox): `{baseDir}/venv` in this repo is a **macOS** venv (Mach-O) and will not run inside the Linux Docker sandbox.
> OpenClaw will create a Linux venv at `{baseDir}/.venv` inside the sandbox and install deps from `{baseDir}/requirements.txt`.

```bash
{baseDir}/.venv/bin/python3 {baseDir}/scripts/query.py \
  --data-dir {baseDir}/data \
  --query "你的問題" \
  --embedding-model qwen3-embedding \
  --llm-model qwen3:8b
```

**Output Format for Agents:**
When responding to user queries:
1. State estimated completion time (typically 5-10 seconds)
2. Execute query and return results directly
3. Do NOT describe the process or show progress messages
4. Include Wiki links and similarity scores in results

## Examples

- "怎樣賺最多錢?"
- "Elliott 喜歡什麼禮物?"
- "漁場怎樣升級?"
- "哪些作物收益最高?"
- "怎樣快速提升人氣度?"
- "能源核心在哪裡找?"

## Requirements

- Python 3.8+
- **Ollama with models** (必須已下載):
  - `qwen3-embedding` (4.7 GB, for embeddings)
  - `qwen3:8b` (5.2 GB, for LLM inference)
- Python venv 已配置完成,所有依賴已安裝(numpy, faiss-cpu, ollama, requests, beautifulsoup4)
- 30 GB 磁盤空間(模型 + 爬蟲結果 + 索引)
- 8-10 GB RAM (qwen3:8b is lighter than GLM-4.7-Flash)

## How It Works

1. **Crawl** — BFS 爬取 zh.stardewvalleywiki.com(100+ 頁)
   - 動態發現:從首頁自動提取所有內容連結
   - 去重機制:URL 規範化 + MD5 hash 防重複

2. **Chunk** — 將每頁分成 500 字的段落

3. **Embed** — 用 qwen3-embedding 生成 4096 維向量
   - 100 頁 → 1063+ embeddings chunks
   - 總大小:~17 MB

4. **Index** — 用 FAISS 建立本機索引
   - 秒級相似度搜尋
   - 無需網路或 API

5. **Query** — 用戶提問 → 相似度搜尋 → qwen3:8b 生成答案
   - 返回可溯源的 Wiki 連結和相似度分數

## Performance

- Query latency: 5-10 seconds(使用 qwen3:8b,比 glm-4.7-flash 快 5-10 倍)
  - Embedding generation: <1 sec
  - FAISS search: ~0.1 sec
  - LLM generation: ~5-8 sec
- Throughput: ~6-10 queries/minute
- Memory: ~8-10 GB
- Storage: ~17 MB (indexed data)

## Troubleshooting

- **"Ollama connection failed"** → Make sure `ollama serve` is running
- **"Model not found"** → Download with `ollama pull qwen3:8b`
- **"No embeddings found"** → Run setup first
- **"Slow response"** → Check if Ollama is overloaded; try reducing --top-k

## Update Data

To refresh Wiki data:

```bash
python3 {baseDir}/scripts/crawl_wiki.py --max-pages 100 --output {baseDir}/data
python3 {baseDir}/scripts/build_vectors.py --input {baseDir}/data/raw_pages.json --output {baseDir}/data
```

Related Skills

ceo-advisor

16
from diegosouzapw/awesome-omni-skill

Executive leadership guidance for strategic decision-making, organizational development, and stakeholder management. Includes strategy analyzer, financial scenario modeling, board governance frameworks, and investor relations playbooks. Use when planning strategy, preparing board presentations, managing investors, developing organizational culture, making executive decisions, or when user mentions CEO, strategic planning, board meetings, investor updates, organizational leadership, or executive strategy.

backend-expert-advisor

16
from diegosouzapw/awesome-omni-skill

Backend expert guidance for API/DB/Security/Architecture

wikidata-search

16
from diegosouzapw/awesome-omni-skill

Search for items and properties on Wikidata and retrieve entity details, claims, and external identifiers. Supports both keyword search (Wikidata Action API) and semantic/hybrid search (Wikidata Vector Database), plus direct entity retrieval (Special:EntityData) and structured querying (WDQS SPARQL).

legal-advisor

16
from diegosouzapw/awesome-omni-skill

Draft privacy policies, terms of service, disclaimers, and legal notices. Creates GDPR-compliant texts, cookie policies, and data processing agreements.

codex-advisor

16
from diegosouzapw/awesome-omni-skill

Get a second opinion from OpenAI Codex CLI for plan reviews, code reviews, architecture decisions, and hard problems. Use when you need external validation, want to compare approaches, or are stuck on a difficult problem.

architecture-advisor

16
from diegosouzapw/awesome-omni-skill

Helps solo developers with AI agents choose optimal architecture (monolithic/microservices/hybrid)

advisor

16
from diegosouzapw/awesome-omni-skill

Interactive workflow advisor that helps you choose optimal AI primitives from agentconfig.org based on your specific workflow needs, skill level, and tooling preferences. Use when deciding which primitives to implement or how to structure your AI configuration.

Advisory Board Builder

16
from diegosouzapw/awesome-omni-skill

Recruit, structure, and manage advisory boards for strategic guidance

tech-advisor

16
from diegosouzapw/awesome-omni-skill

Recomienda stack tecnológico óptimo basado en requisitos del proyecto

latency-advisor

16
from diegosouzapw/awesome-omni-skill

Provides SRE latency optimization advice for Claude API usage. Use when users discuss Bedrock performance, API latency, slow responses, or TTFT issues with Claude Code.

boardroom-advisor

16
from diegosouzapw/awesome-omni-skill

Consult a virtual board of 4 strategic advisors (Donald Miller, Seth Godin, Alex Hormozi, Daniel Priestley) on any major business decision. Two rounds of argument + rebuttal, then a decision brief, interactive dashboard, and clear recommendation.

advisor-triggers

16
from diegosouzapw/awesome-omni-skill

Detects when user requests warrant critical analysis via /advise command