import - Fetch URL (via Playwright if blocked) or copy local path to docs/
Imports documentation from a URL or local path into a structured cache, using Context7, WebFetch, or Playwright to fetch web content or copying local files.
When to use
Use this when you need to fetch and save external documentation or local files for later reference, ensuring versioned organization and bypassing web blocks.
How to invoke
manual via /import:url-or-path, auto-invoked when relevant
npx claudepluginhub marcel-bich/marcel-bich-claude-marketplaceYou are executing the /import:url-or-path command. Import documentation from a URL or local path.
$ARGUMENTS - URL (https://...) or local file/directory pathAll imports go to mabi-import/ with intelligent substructure:
mabi-import/{library}/{version}/
datahub.md
data-importer.md
Example: https://docs.pimcore.com/platform/Data_Importer/ (version 2025.4)
-> mabi-import/pimcore/v2025.4/data-importer.md
mabi-import/website/{domain}/
{page-name}.md
Example: https://example.com/api/docs
-> mabi-import/website/example.com/api-docs.md
mabi-import/local/{project-or-context}/
{filename}
Determine context from:
Example: /home/user/projects/myapp/docs/config.md
-> mabi-import/local/myapp/config.md
Check if $ARGUMENTS is:
http:// or https://If local path:
Check if exists:
ls -la "$ARGUMENTS"
Determine context for folder structure:
/projects/myapp/ -> myapp)Confirm with user:
You specified a local path: {path}
Suggested location: mabi-import/local/{context}/{filename}
Is this correct? (The path will be copied, not moved)
Use AskUserQuestion to confirm or let user provide different context.
If confirmed:
mkdir -p "mabi-import/local/{context}"
cp -r "$ARGUMENTS" "mabi-import/local/{context}/"
If URL, try methods in this order. Inform user at each step:
mcp__plugin_context7_context7__resolve-library-id to check if library existsmcp__plugin_context7_context7__query-docs to get relevant contentmabi-import/{library}/{version}/mabi-import/{library}/{version}/mabi-import/website/{domain}/browser_navigate to the URLbrowser_snapshot to get page contentIMPORTANT: Do NOT use headless mode. Many sites (especially those using Cloudflare) detect and block headless browsers. The visible browser window is intentional - it allows passing bot detection and manual captcha solving if needed.
Try to detect version from:
/v2/, /v2025.4/, /version/19/IMPORTANT: Never use "latest" as version folder name. If no version can be detected:
Exception: For website/ and local/ categories, version is not required (content is not versioned).
Target location based on type:
mabi-import/{library}/{version}/{page-name}.mdmabi-import/website/{domain}/{page-name}.mdmabi-import/local/{context}/{filename}Format saved file:
---
source: {original URL or path}
imported: {timestamp}
method: {context7|webfetch|playwright|local}
version: {exact version - NEVER "latest"} # omit for local/website
---
# {Title}
{Content}
Note: version field is only required for library documentation. Omit it for local/ and website/ imports.
On success:
Documentation imported:
Source: {url or path}
Method: {context7|webfetch|playwright|local}
Version: {version}
Saved: mabi-import/{structure}/{name}.md
Size: {file size}
Use /import:list to see all cached docs.
Use /import:search <query> to search within docs.
On complete failure:
Failed to import documentation:
Source: {url or path}
Tried:
1. Context7: {not found / error}
2. WebFetch: {403 blocked / error}
3. Playwright: {error}
Cannot fetch this URL automatically.
Suggestions:
- Check if URL is correct and accessible
- Copy content manually and use /import:url-or-path /local/path
- Check if Context7/Playwright plugins are installed