Search Interceptor

소개

Search Interceptor is a performance-optimization skill for Claude Code that monitors file-system activity to prevent inefficient bulk reading patterns. By identifying 'tool familiarity bias'—where an AI defaults to sequential file reads or broad glob patterns—it redirects the workflow toward semantic search via claudemem. This ensures more relevant code context is retrieved with significantly lower token consumption, making large-scale codebase investigations faster, cheaper, and more precise.

주요 기능

• Automatic interception of 3+ sequential file reads
• Token usage optimization for large-scale investigations
• Real-time evaluation of investigative intent versus tool efficiency
• Broad glob pattern validation and redirection
• Semantic search integration via claudemem
• 1 GitHub stars

사용 사례

• Auditing complex API integrations across multiple nested directories
• Investigating security flows like authentication without reading every related file
• Discovering architectural patterns using semantic concepts instead of rigid filename patterns