Links Integration
Overview
Web Intelligence integrates with the Links MCP server for URL bookmarking and management. This integration enables discovered links to be persisted with metadata for future reference.
Links MCP Capabilities
Available Tools
| Tool | Purpose |
|---|---|
links.save |
Save single URL with metadata |
links.save_batch |
Batch save multiple URLs |
links.get |
Retrieve saved link by ID |
links.search |
Search links by query |
links.list |
List all links with filtering |
links.systems |
List available categorization systems |
links.delete |
Remove a saved link |
Auto-Features
- URL Normalization: Automatic URL cleaning and standardization
- Domain Categorization: Links are categorized by domain
- Metadata Fetching: Title and description auto-extracted
- Hash Deduplication: MD5 hash prevents duplicate URLs
Web Intelligence Integration Tools
web.discover_links
web.discover_links(url: str, depth: int = 1) -> List[str]
Uses spider_rs get_links() to extract all links from a page. Returns list of discovered URLs.
web.save_link
web.save_link(url: str, system: str = None, tags: List[str] = None) -> dict
Wrapper for links.save() with optional system categorization and tagging.
web.save_discovered
web.save_discovered(links: List[str], system: str = None) -> dict
Batch save discovered links. Calls links.save_batch() for efficiency.
web.filter_links
web.filter_links(links: List[str], criteria: dict) -> List[str]
Pre-filter links before saving. Criteria can include:
- domains: List of allowed domains
- exclude_domains: List of blocked domains
- patterns: Regex patterns to match
- exclude_patterns: Regex patterns to exclude
Data Flow
spider_rs.get_links() → web.filter_links() → web.save_discovered() → links.save_batch()
Deduplication
Handled automatically by Links MCP via MD5 hash of normalized URL. No need for Web Intelligence to track duplicates.
Related Documentation
- Links MCP Server: Ports 6635/6636
- Spider Engine Reference: KB node 206d2549