vincentkoc

vincentkoc / tokenjuice

Public

🧃 Token weight loss. Lean output compaction for terminal-heavy agent workflows. Works as a native CLI tool or as an extension to popular coding and agent frameworks.

openclaw-plugin optimization token tui-app
14
1
100% credibility
Found Apr 16, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Tokenjuice cleans up verbose terminal command outputs for AI coding agents by summarizing key details while storing full logs if needed.

How It Works

1
😩 Messy Outputs Frustrate You

You're using an AI coding helper, but long, noisy command results like file lists or test logs overwhelm it and waste time.

2
🔍 Discover Tokenjuice

You learn about Tokenjuice, a simple helper that cleans up those bulky outputs so your AI gets just the key info.

3
📦 Add It Easily

With one quick step, you add Tokenjuice to your computer, ready to tidy any command's output.

4
Connect to Your AI Helper
🤖
For Claude Code

Tell it to connect Claude, and it sets up the magic link.

🧠
For Codex

Or connect Codex, and outputs get slimmed instantly.

5
✨ Run Cleaner Commands

Now, wrap any command, and Tokenjuice delivers short, smart summaries—your AI understands perfectly without the clutter!

6
📊 Check Savings Anytime

Peek at stats or saved full outputs to see how much space and time you've saved.

🎉 AI Workflows Fly

Your AI helper responds faster and smarter, turning chaotic terminals into smooth, efficient coding magic.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is tokenjuice?

Tokenjuice is a TypeScript CLI for compacting verbose terminal output in agent workflows, trimming noise from commands like git status, pnpm test, or docker-compose up before feeding it to LLMs. It runs tools unchanged, applies JSON rules for token weighting, and returns lean summaries—cutting token github push walls or token github clone logs to essentials. Use it standalone via tokenjuice wrap or as hooks for Claude Code and Codex CLI.

Why is it gaining traction?

Its rule-based compaction is deterministic and extensible, with user/project overrides for custom token weighting in agent chains, plus artifact storage for raw outputs. Escape hatches like --raw bypass and doctor hooks prevent drift, while stats/discover commands spotlight weak spots like poor token github repo reductions. No framework lock-in means it slots into any CLI-heavy agent setup.

Who should use this?

AI agent builders chaining terminal tools into LLM prompts, especially those hitting context limits on git/pnpm/docker noise. Devs extending Claude Code or Codex CLI for token-efficient coding agents. Terminal scripters in devops wanting token weighting for long-range language modeling without manual parsing.

Verdict

Worth a spin for agent prototypes—CLI commands and hooks deliver immediate token savings. At 14 stars and 1.0% credibility, it's pre-traction but mature enough with strong docs, tests, and diagnostics for side projects.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.