file-read

name	file-read
description	Read files efficiently to minimize token usage — use targeted reads instead of dumping entire files

file-read

Read files efficiently — fetch only what you need. Never cat a large file when a subset suffices.

Decision Guide

Situation	Command
Know what you're looking for	grep -n "symbol" file
Need first N lines	head -n 50 file
Need last N lines	tail -n 30 file
Need lines M–N	sed -n 'M,Np' file
Unknown file, check size first	wc -l file && head -n 30 file
File < 100 lines	cat file

Patterns

wc -l file.txt                                     # check size before reading

grep -n "pattern" file.txt                         # find lines + line numbers
grep -n -A 5 -B 2 "pattern" file.txt              # with surrounding context
grep "^import\|^from" file.py                      # filter specific line types
grep -v "^#\|^$" file.txt                          # skip comments/blank lines

head -n 50 file.txt                                # first 50 lines
tail -n 30 file.txt                                # last 30 lines
sed -n '10,30p' file.txt                           # lines 10–30

sed -n '/^## Section/,/^## /p' file.txt            # between markers

grep -n "^def my_func" file.py                     # find block start line
sed -n '42,60p' file.py                            # then extract the block

{ sed -n '1,10p' f.txt; echo "---"; sed -n '80,90p' f.txt; }  # non-contiguous

wc -l file.txt && head -n 5 file.txt && echo "..." && tail -n 5 file.txt  # summarize

Rules

• Check wc -l before reading unknown files
• Use grep first when you know what you're looking for
• Read in chunks (30–50 lines); fetch more only if needed
• Avoid cat on files > 200 lines unless full content is required