Changes

• Updated RipgrepSearch.parse() to properly handle JSON output structure
• Now accumulates content from 'begin' to 'end' for each file
• Includes both match and context lines in code snippets

Why

The Scout agent was unable to retrieve code snippets because the parsing logic only extracted 'match' types, missing context lines and not accumulating content properly.

Fixes the issue where Engineer agents had limited context due to incomplete code snippet extraction.