Search documentation content with a regex pattern. Like the Unix ‘grep’ command, but for indexed documentation. Searches through all pages and returns matches with context.
Supports asymmetric context lines (A for after, B for before), case sensitivity, whole word matching, and multiple output modes.
API key must be provided in the Authorization header
Flexible data source identifier (UUID, display name, or URL)
Regex pattern to search for
"authentication.*token"
Limit search to this virtual path prefix
Lines before AND after each match (shorthand for A/B). Overridden by A or B if specified.
0 <= x <= 10Lines after each match (like grep -A). Overrides context_lines for after.
0 <= x <= 20Lines before each match (like grep -B). Overrides context_lines for before.
0 <= x <= 20Case-sensitive matching (default is case-insensitive)
Match whole words only
Treat pattern as literal string, not regex
Maximum matches to return per file
1 <= x <= 100Maximum total matches to return
1 <= x <= 1000Output format:
content, files_with_matches, count Add >>markers<< around matched text in results
Include line numbers in results
Group matches by file in results
Search completed successfully
Matches grouped by file (when output_mode is 'content')
List of file paths (when output_mode is 'files_with_matches')
Match counts per file (when output_mode is 'count')
The pattern that was searched
Path filter that was applied
Number of files that contained matches
Whether results were truncated due to limits
Applied search options