# Token Efficiency

> Use to optimize context usage during file inspection, debugging, and targeted testing.

- Skill: `sahilkhan30/token-efficiency` (Agent Skill)
- Install (CLI): `npx skillmds@latest add sahilkhan30/token-efficiency`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sahilkhan30/token-efficiency/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: SahilKhan30 (https://skillmd.com/u/sahilkhan30)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/sahilkhan30/token-efficiency

---


# Token Efficiency Skill

## Purpose
Maximize context efficiency during multi-agent collaboration by avoiding unnecessary broad inspections and redundant file reads.

---

## 1. Directives

### RECOMMENDED
- **Search Before Read**: Use targeted grep/symbol searches before loading full file contents.
- **Avoid Unnecessary Inspection**: Inspect enough context to make a correct decision, but do not read unrelated packages.
- **Do Not Re-Read Unchanged Files**: Reuse memory of unchanged files from previous context steps.
- **Targeted Test Execution**: Run single unit/integration tests (e.g. `mvn test -Dtest=TransactionServiceTest`) rather than entire suites during iterative build loops.