packet28

Your AI agent reads
too much. We fix that.

AI coding agents waste tokens reading entire files. Packet28 compresses what your agent sees — a 9,243-line file becomes ~40 relevant lines. Result: 164x fewer tokens on large codebases.

Works with: Claude Code · Cursor · Codex · Copilot · Windsurf

$ npm install -g packet28

1 get started 2 source 3 npm 4 docs

What it does

READCompresses file reads

Your agent tries to read a 9,243-line file. Packet28 intercepts it and returns only the ~40 lines that matter for the current task. No wasted tokens on irrelevant code.

MAPMaps your repo

Builds a structural map of your codebase — files, symbols, dependencies — so the agent navigates directly to relevant code instead of exploring blindly.

MEMRemembers across sessions

A background daemon persists context between agent restarts. Pick up where you left off without the agent re-reading your entire project from scratch.

SETOne command setup

Auto-detects Claude Code, Cursor, Codex, Copilot, or Windsurf and configures everything. One command: packet28 setup --root .

The numbers

Real benchmark on a 12MB Java codebase (534 files).

WITHOUT PACKET28

139,000 tokens

WITH PACKET28

849

164x

fewer tokens consumed

Task: "Understand StringUtils.wrap and fix null delimiter handling"
Naive agent reads entire 9,243-line files. Packet28 serves ~40 relevant lines + anchors.

Up and running in 30 seconds

—

□

terminal

# install globally

$ npm install -g packet28

# set up hooks + MCP in your project

$ packet28 setup --root .

✓ Detected runtimes: Claude Code, Cursor

✓ Hooks installed. MCP servers configured.

# verify everything works

$ packet28 doctor --root .

✓ All checks passed.

Your AI agent readstoo much. We fix that.

What it does

The numbers

Up and running in 30 seconds

Built by

Your AI agent reads
too much. We fix that.