L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Déploiement et opérations · SKILL.md

rtk-token-optimization

rtk-token-optimization

High-performance CLI proxy that reduces LLM token consumption by 60-90%. Transparently rewrites 100+ common dev commands (git, gh, cargo, npm, docker, kubectl, aws, etc.) via PreToolUse hook. Single Rust binary, zero dependencies.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: rtk-token-optimization description: > High-performance CLI proxy that reduces LLM token consumption by 60-90%. Transparently rewrites 100+ common dev commands (git, gh, cargo, npm, docker, kubectl, aws, etc.) via PreToolUse hook. Single Rust binary, zero dependencies. source: https://github.com/rtk-ai/rtk license: Apache-2.0 trust_tier: B-proven-community

Purpose

Reduce LLM token consumption in Claude Code sessions by 60-90% without changing developer workflow. All Bash commands are transparently rewritten to their rtk equivalents via a PreToolUse hook.

How It Works

Without rtk:                              With rtk:
Claude --git status--> shell --> git      Claude --git status--> RTK --> git
  ^        ~2,000 tokens (raw)    |        ^    ~200 tokens     |       |
  +-------------------------------+        +---- (filtered) ----+-------+

Four strategies applied per command type:

  1. Smart Filtering - Removes noise (comments, whitespace, boilerplate)
  2. Grouping - Aggregates similar items (files by directory, errors by type)
  3. Truncation - Keeps relevant context, cuts redundancy
  4. Deduplication - Collapses repeated log lines with counts

Supported Commands

CategoryCommandsTypical Savings
Filesls, read, find, grep, diff70-80%
Gitstatus, diff, log, add, commit, push75-92%
GitHub CLIpr list, issue list, run list80%
Test Runnerscargo test, npm test, pytest, go test, vitest, playwright90%
Build & Lintcargo build, tsc, next build, eslint, ruff, clippy80-85%
Package Managerspnpm list, pip list, bundle install80%
Containersdocker ps/logs, kubectl pods/logs80%
AWSsts, ec2, lambda, logs, s3, cloudformation80%
Datajson, env, log, curl, wget70-80%

Token Savings (30-min Session Estimate)

OperationFrequencyStandardrtkSavings
ls / tree10x2,000400-80%
cat / read20x40,00012,000-70%
grep / rg8x16,0003,200-80%
git status10x3,000600-80%
git diff5x10,0002,500-75%
cargo test / npm test5x25,0002,500-90%
Total~118,000~23,900-80%

Usage

After installation, rtk works automatically via the PreToolUse hook. No manual changes needed.

Analytics Commands

rtk gain                # Summary stats
rtk gain --graph        # ASCII graph (last 30 days)
rtk gain --daily        # Day-by-day breakdown
rtk discover            # Find missed savings opportunities
rtk session             # Show RTK adoption across recent sessions

Manual Usage

rtk git status          # Compact git status
rtk read file.rs        # Smart file reading
rtk test cargo test     # Show failures only
rtk lint                # ESLint grouped by rule/file
rtk docker ps           # Compact container list

Prerequisites

  • rtk binary: brew install rtk or curl -fsSL https://raw.githubusercontent.com/rtk-ai/rtk/refs/heads/master/install.sh | sh
  • jq: required by the hook script

Integration

  • Hook: hooks/rtk-rewrite.sh (PreToolUse:Bash, last in chain)
  • Config: ~/.config/rtk/config.toml for customization
  • Tee recovery: failed command full output saved to ~/.local/share/rtk/tee/

References

Skills enregistrées