dneprokos/educational-resource-searcher

Educational Resource Searcher

This skill searches popular educational platforms for high-quality learning resources based on user-specified topics, with optional filters for language, platform, free/paid status, and maximum results.

When this skill fits

• User asks to find tutorials, courses, videos, or learning materials on a specific topic (e.g., "SQL basics", "machine learning").
• Requests include filters like language ("in Ukrainian"), platform ("on YouTube"), free/paid ("free videos"), or max results.
• Queries about learning resources, educational content, or skill-building materials.
• Do NOT use for general web searches, non-educational content, or when no topic is specified.

Workflow

1. Parse the user's query to extract parameters:

   • topics: Required, at least one (e.g., "SQL database basics").
   • language: Optional, default "english".
   • platforms: Optional, list from references/resources.md; default all.
   • max_items: Optional, default 5.
   • free_only: Optional, boolean; default false (include paid).

2. Validate: If no topics provided, return error message: "You need to specify at least one topic."

3. Load platform details from references/resources.md.

4. For each selected platform, check its Fetch Strategy in references/resources.md and apply the matching approach:

   • WebFetch platforms (Coursera, W3Schools): Construct the search URL from the template and fetch the page directly using the WebFetch tool. Extract results from the returned HTML.
   • WebSearch platforms (YouTube, Udemy, LinkedIn Learning, Pluralsight, edX, Khan Academy): Use the WebSearch tool with the query pattern site:{platform_domain} {topic} {language} course tutorial. Extract titles, links, ratings, and metadata from the search result snippets returned.
   • For both strategies: collect title, link, rating (if available), views/enrollments, free/paid status. Filter results based on free_only and language (if detectable).

5. Aggregate results from all platforms.

6. Sort by rating descending, then views/enrollments descending.

7. Limit to max_items.

8. Output in ASCII table format.

Hard rules

• Always require at least one topic; return error otherwise.
• Default language: English.
• Default max_items: 5.
• Default platforms: All listed in references.
• Prioritize highest rated and most viewed resources.
• Only include resources with valid links.
• If no results found, output: "No educational resources found for the specified criteria."

Output format

Return a Markdown table:

| Platform | Title | Link | Rating | Views | Free/Paid | | -------- | ------------------- | ----------- | ------ | ----- | --------- | | YouTube | SQL Basics Tutorial | https://... | 4.8 | 1.2M | Free | | ... | ... | ... | ... | ... | ... |

Examples

Input: "find sql database basics on youtube in ukrainian language free video"

Output: | Platform | Title | Link | Rating | Views | Free/Paid | |----------|-------|------|--------|-------|-----------| | YouTube | SQL Basics in Ukrainian | https://youtube.com/watch?v=abc | 4.5 | 50K | Free |

Input: "tutorials on python programming, free only, max 3"

Output: (Table with up to 3 free Python resources from various platforms)

Fetch strategy notes

Many educational platforms render search results client-side via JavaScript. The WebFetch tool only retrieves the initial static HTML, so it returns an empty shell with no results for these sites. The skill uses two strategies:

• WebFetch: For platforms that server-render their search results (currently Coursera and W3Schools). Faster and returns structured HTML.
• WebSearch: For JS-heavy platforms (YouTube, Udemy, LinkedIn Learning, Pluralsight, edX, Khan Academy). Uses a search engine query with a site: filter to retrieve indexed, pre-rendered results from snippets. Less structured than parsed HTML but reliable across all browsers and auth walls.

If a WebFetch call returns only navigation and footer content with no actual results, the platform is JS-rendered and must be switched to the WebSearch strategy.

Extending this skill

To add new platforms, update references/resources.md with the new entry following the existing format. Always include:

1. The Fetch Strategy (WebFetch or WebSearch) — test by attempting a WebFetch on the platform's search URL. If the result contains no course/video listings, the platform requires WebSearch.
2. The Site domain (for WebSearch platforms), used in the site: filter query.
3. The Search URL template (kept for reference even for WebSearch platforms).

The skill will automatically include the new platform in all searches once added.