v0.3.0

• added support for auto-prompt caching to reduce api costs
• fixed search result deduplication across multiple queries
• new crawl_depth parameter for recursive page crawling
• improved error handling for rate limit responses
• refactored internal request batching logic

v0.2.0

• added support for crawling multiple pages with configurable depth limits
• improved error handling for rate limit and network failures
• refactored search result parsing to support additional content types
• fixed bug where search dates were not properly formatted in requests
• updated dependencies to latest exa-py sdk version
• added request timeout configuration option

v0.1.0

• initial release of exa-mcp server with search and crawling capabilities
• implements mcp protocol for exa api integration
• adds search tool with query, date range, and result filtering options
• adds contents tool for retrieving full page content by url
• supports both text and highlight extraction modes
• includes basic error handling and request validation