Wikipedia SearchSAFE

The largest open-source medical AI skills library for OpenClaw🦞.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: wikipedia-search
description: Search and fetch structured content from Wikipedia using the MediaWiki API for reliable, encyclopedic information
metadata:
  openclaw:
    emoji: "📚"
    requires:
      bins: ["python3"]
---

# Wikipedia Search Skill

This skill enables Claude to search and fetch content from Wikipedia using the MediaWiki API. It provides more reliable, structured access to Wikipedia than general web search, with support for summaries, full article content, and multi-language access.

## When to Use

Use this skill when you need:

- **Encyclopedic knowledge** — Factual information about people, places, events, concepts
- **Historical information** — Well-documented historical facts, dates, and events
- **Scientific and technical concepts** — Definitions, explanations, and overviews
- **Biographical information** — Details about notable people
- **Geographic information** — Countries, cities, landmarks, and their details
- **Quick fact verification** — Cross-checking encyclopedic facts
- **Structured content** — Need article sections, categories, and links
- **Multi-language content** — Access Wikipedia in languages other than English

## When NOT to Use

Avoid using this skill for:

- **Current events or breaking news** — Wikipedia may not be up-to-date for very recent events (use web-search instead)
- **Controversial or disputed topics** — Wikipedia can have bias or edit wars; use multiple sources
- **Opinions or analysis** — Wikipedia aims for neutral point of view, not opinion
- **How-to guides or tutorials** — Wikipedia is encyclopedic, not instructional (look for specialized resources)
- **Code examples or technical documentation** — Use official docs or specialized sites
- **Personal or local information** — Wikipedia only covers notable topics
- **Real-time data** — Stock prices, weather, sports scores (use web-search instead)

## Usage

The skill is invoked by running the wiki script with a query:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "your query" [options]
```

### Options

- `--mode [search|summary|full]` — Operation mode (default: summary)
  - `search`: Search for page titles matching the query
  - `summary`: Get a concise summary of a specific page
  - `full`: Get the complete article with sections and structure
- `--sentences N` — Number of sentences for summary mode (default: 5)
- `--lang CODE` — Language code (default: en)

### Common Language Codes

- `en` — English
- `es` — Spanish
- `fr` — French
- `de` — German
- `it` — Italian
- `ja` — Japanese
- `zh` — Chinese
- `ru` — Russian
- `ar` — Arabic
- `pt` — Portuguese

## Examples

### 1. Search for Page Titles

Search for Wikipedia articles matching a term:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "quantum computing" --mode search
```

Output:
```json
{
  "mode": "search",
  "query": "quantum computing",
  "results": [
    {
      "title": "Quantum computing",
      "description": "Study of computers based on quantum phenomena",
      "url": "https://en.wikipedia.org/wiki/Quantum_computing"
    },
    ...
  ],
  "count": 10
}
```

### 2. Get a Summary

Get a concise summary of a Wikipedia page:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "Python (programming language)" --mode summary
```

Or with custom sentence count:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "Albert Einstein" --mode summary --sentences 3
```

Output:
```json
{
  "mode": "summary",
  "title": "Albert Einstein",
  "exists": true,
  "url": "https://en.wikipedia.org/wiki/Albert_Einstein",
  "summary": "Albert Einstein was a German-born theoretical physicist...",
  "categories": ["1879 births", "1955 deaths", "German physicists", ...]
}
```

### 3. Get Full Article Content

Retrieve the complete article with sections:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "Machine learning" --mode full
```

Output:
```json
{
  "mode": "full",
  "title": "Machine learning",
  "exists": true,
  "url": "https://en.wikipedia.org/wiki/Machine_learning",
  "summary": "Machine learning is a field of study in artificial intelligence...",
  "sections": [
    {
      "title": "History and relationships to other fields",
      "level": 0,
      "text": "..."
    },
    ...
  ],
  "categories": [...],
  "links": ["Artificial intelligence", "Neural network", ...]
}
```

### 4. Search in Different Languages

Search or fetch content in other languages:

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "Paris" --mode summary --lang fr
```

```bash
python ~/.openclaw/skills/wikipedia-search/scripts/wiki.py "東京" --mode summary --lang ja
```

## Workflow

When using this skill, follow this workflow:

### For Unknown Topics (Search First)

1. **Search for titles** — Use `--mode search` to find relevant articles
2. **Choose the right article** — Pick the most relevant title from results
3. **Get summary or full content** — Fetch the article with `--mode summary` or `--mode full`
4. **Parse JSON output** — Extract relevant information from structured data
5. **Cite the source** — Include Wikipedia URL when presenting information

### For Known Topics (Direct Fetch)

1. **Use exact or close title** — Wikipedia is good at matching similar titles
2. **Fetch summary** — Use `--mode summary` for quick facts
3. **Get full content if needed** — Use `--mode full` for comprehensive information
4. **Extract sections** — Parse specific sections for focused information

### JSON Output Structure

#### Search Mode
```json
{
  "mode": "search",
  "query": "original query",
  "results": [
    {
      "title": "Page Title",
      "description": "Brief description",
      "url": "https://en.wikipedia.org/wiki/Page_Title"
    }
  ],
  "count": 10
}
```

#### Summary Mode
```json
{
  "mode": "summary",
  "title": "Article Title",
  "exists": true,
  "url": "https://en.wikipedia.org/wiki/Article_Title",
  "summary": "Summary text...",
  "c

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

declared (2 observation(s))

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.