drupal-docs-explorer

Drupal Documentation Explorer

Fetch and return official Drupal documentation relevant to a query. This skill searches a curated catalog of essential drupal.org and api.drupal.org URLs, identifies the most relevant pages, fetches their full content, and returns it for the calling agent to use.

Arguments

• $1 — Query (required). A natural language description of what you need to know. Examples:
  • "entity API field definitions"
  • "how to create a custom module"
  • "form API validation and submission"
  • "routing and controllers"
  • "services and dependency injection"
  • "configuration management export import"
  • "caching strategies and cache tags"
  • "Twig theming and templates"
  • "hook system event subscribers"
  • "automated testing PHPUnit"

Execution

You are a fast research agent. Your job is to find and fetch documentation, not to analyze or act on it. Return the raw content so the calling agent can use it.

Step 1: Load the URL catalog

Read the URL catalog file at references/drupal-docs-urls.md (in this skill's directory). This contains all curated Drupal documentation URLs organized by category, each with a brief description of its content.

Step 2: Match URLs to the query

From the catalog, identify all pages relevant to the query. Match by:

• URL path segments (e.g., query "entity API" matches URLs containing /entity-api, /entity, /content-entities)
• Description keywords (e.g., query "dependency injection" matches entries described as "Services and dependency injection container")
• Semantic relevance (e.g., query "custom module" matches /creating-modules, /module-file, /info-yml, /hooks, /routing-system)
• Category context (e.g., query "testing" should match entries under the Testing category, plus related entries like coding standards)

Be generous with matching — it is better to fetch an extra page than to miss a relevant one. But stay focused: do not fetch unrelated categories unless the query is broad.

Matching guidelines:

• For API queries, include both the overview page and the specific API page
• For "how to" queries, include both tutorial/guide pages and the relevant API reference pages
• For troubleshooting queries, include security pages, change records, and relevant API docs
• Always include the most specific match first, then broaden to related topics

Step 3: Fetch matched pages

For each matched URL, fetch the URL content to retrieve the page. Fetch pages in parallel where possible for speed.

When fetching, extract the complete content of each page. Return all text, code examples, tables, steps, warnings, and notes. Do not summarize — return everything.

If a URL redirects, follow the redirect and note the final URL in the output.

Step 4: Return results

Return all fetched content organized by page, with clear headers:

## <Page Title> — <URL>

<full page content>

---

## <Next Page Title> — <URL>

<full page content>

If a page fails to load (404, 503, timeout), note the failure and continue with remaining pages. Suggest the user check whether the URL has moved by searching drupal.org directly.

Tips for effective queries

• Be specific: "Form API AJAX callbacks" will yield better results than just "forms"
• Use Drupal terminology: "render arrays" not "page rendering", "entity bundles" not "content types API"
• Combine concepts: "routing access control" will match both routing and permissions pages
• Ask about versions: If you need version-specific docs, mention "Drupal 10" or "Drupal 11" in your query

Catalog maintenance

The URL catalog at references/drupal-docs-urls.md is a manually curated list. To maintain it:

• Periodically verify URLs still resolve (drupal.org restructures docs occasionally)
• Add new URLs when major documentation pages are published
• Remove URLs that consistently 404 or redirect to unrelated content
• When Drupal releases a new major version, update api.drupal.org URLs to reference the latest version