Tools Overview¶
The LinkDing MCP Server provides 9 comprehensive tools for managing your bookmarks. Each tool is designed to work seamlessly with LLMs and follows MCP best practices.
Available Tools¶
| Tool | Category | Description |
|---|---|---|
search_bookmarks |
Search | Search bookmarks with flexible filters |
add_bookmark |
Management | Add new bookmarks with metadata |
get_bookmark |
Retrieval | Get specific bookmark by ID |
update_bookmark |
Management | Update existing bookmark properties |
delete_bookmark |
Management | Delete bookmarks permanently |
archive_bookmark |
Organization | Archive bookmarks |
unarchive_bookmark |
Organization | Unarchive bookmarks |
check_url |
Utility | Check if URL is already bookmarked |
list_tags |
Tags | List all available tags |
list_bookmarks_by_tag |
Tags | Filter bookmarks by specific tag |
Tool Categories¶
🔍 Search Tools¶
search_bookmarks: Primary search interface with multiple filter optionslist_bookmarks_by_tag: Tag-specific filtering
📚 Bookmark Management¶
add_bookmark: Create new bookmarksget_bookmark: Retrieve bookmark detailsupdate_bookmark: Modify existing bookmarksdelete_bookmark: Remove bookmarks
📦 Organization Tools¶
archive_bookmark: Archive for later referenceunarchive_bookmark: Restore from archive
🏷️ Tag Management¶
list_tags: Discover available tags
🔧 Utility Tools¶
check_url: Prevent duplicates and get metadata
Common Usage Patterns¶
Research Workflow¶
# 1. Check if already bookmarked
check_url(url="https://research-paper.com")
# 2. Add if new
add_bookmark(
url="https://research-paper.com",
tags=["research", "ai"],
notes="Important findings on neural networks"
)
# 3. Find related content
search_bookmarks(tag="research")
Content Organization¶
# 1. Review unread items
search_bookmarks(unread_only=True)
# 2. Update with proper tags
update_bookmark(
bookmark_id=123,
tags=["tutorial", "javascript", "frontend"]
)
# 3. Archive when processed
archive_bookmark(bookmark_id=123)
Discovery and Exploration¶
# 1. See what tags you have
list_tags()
# 2. Explore a category
list_bookmarks_by_tag(tag_name="tutorials")
# 3. Search within category
search_bookmarks(tag="tutorials", query="react")
Error Handling¶
Error handling has a deliberate layer boundary:
LinkDingClientraisesLinkDingError(includingRateLimitError) for expected API, connectivity, and rate-limit failures.- MCP tools catch those domain errors, log them, and return a stable
Error: <message>string because MCP tool responses are values rather than Python exceptions. - Validation and unexpected programming failures are handled at the tool boundary with operation-specific context.
Response Formats¶
Success Responses¶
Tools return structured data in JSON format:
{
"id": 123,
"url": "https://example.com",
"title": "Example Site",
"description": "An example website",
"tags": ["example", "demo"],
"date_added": "2024-01-15T10:30:00Z"
}
Error Responses¶
Errors are returned as descriptive strings:
"Error: Bookmark with ID 999 not found"
"Error: Invalid URL format"
"Error: Connection to LinkDing failed"
Performance Considerations¶
Pagination¶
- Use
limitandoffsetparameters for large result sets - Default limits are set to reasonable values
- Consider memory usage when requesting large datasets
Rate Limiting¶
- The server respects LinkDing's rate limits
- Batch operations when possible
- Use appropriate delays for bulk operations
Caching¶
- LinkDing handles caching internally
- Fresh data is always returned
- No client-side caching implemented
Security Features¶
API Token Protection¶
- Tokens are never logged or exposed
- Secure transmission to LinkDing API
- Environment variable isolation
Input Validation¶
- URL validation for bookmark operations
- Parameter type checking
- SQL injection prevention (handled by LinkDing)
Access Control¶
- Respects LinkDing user permissions
- No privilege escalation possible
- Read-only operations where appropriate
Next Steps¶
Explore detailed documentation for each tool category:
- Search Tools - Advanced search and filtering
- Bookmark Management - CRUD operations
- Tag Management - Organization and categorization
Or jump to specific integration guides:
- Claude Desktop - Desktop integration
- API Reference - Technical details