ashlr__grep
Genome-aware search. Routes through .ashlrcode/genome/ when present; falls back to ripgrep with output truncation.
Server: ashlr-efficiency — servers/efficiency-server.ts
Replaces Claude Code's built-in Grep tool. When a .ashlrcode/genome/ index exists in the project, search is routed through the genome retriever (RAG), which returns pre-extracted knowledge snippets instead of raw file matches — typically ~84 % fewer tokens on warm queries. Without a genome, falls back to ripgrep with output truncation.
Input schema
| Field | Type | Required | Description |
|---|---|---|---|
pattern | string | yes | Regex or literal search pattern |
path | string | no | Directory or file to search in (default: cwd) |
glob | string | no | Glob filter, e.g. **/*.ts |
-i | boolean | no | Case-insensitive match |
output_mode | string | no | content, files_with_matches, or count |
context | number | no | Lines of context around each match |
head_limit | number | no | Max results to return |
Output format
Genome mode returns structured knowledge snippets with source references. Ripgrep fallback returns matched lines with file:line references, truncated to head_limit.
Example
Savings baseline
Genome mode: median −84 % vs raw ripgrep output. Ripgrep fallback: ~40 % via truncation.
Related tools
ashlr__read— read a specific file after locating itashlr__orient— high-level structural searchashlr__glob— find files by name pattern