context-compressor

Name: context-compressor
Availability: InStock
Author: kanno321-create

Compress Claude session context by measuring, slimming, and archiving documents.

Updated Aug 27, 2026

Search Tags:#workflow-automation #documentation-management #context-compression #claude-ai #session-archiving #lost-in-the-middle

Ready to run

npx skills add https://github.com/kanno321-create/shorts_studio --skill context-compressor-kanno321-create

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: context-compressor
Source: https://github.com/kanno321-create/shorts_studio/tree/main/.claude/skills/context-compressor
Command: npx skills add https://github.com/kanno321-create/shorts_studio --skill context-compressor-kanno321-create

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

CLAUDE 세션 시작 시 CLAUDE.md, WORK_HANDOFF.md, SESSION_LOG.md 등이 로딩될 때 과도한 컨텍스트 소비로 인한 Lost-in-the-Middle 위험을 줄이고, 과거 아카이브와 최근 세션의 중복 제거를 통해 초기 컨텍스트를 1,000줄 이내로 유지합니다.

Core Features & Use Cases

과거 아카이브 + 최근 유지 + 중복 제거를 통한 자동 컨텍스트 압축으로 불필요한 텍스트를 줄입니다.
Phase-based 워크플로우로 측정, 다이어트, 압축, 아카이브, 중복 제거를 순차적으로 수행합니다.
최종 산출물로 슬림화된 문서 세트와 아카이브 및 컨텍스트 감사 보고서를 제공합니다.

Quick Start

세션 시작 시 컨텍스트 압축 워크플로우를 실행하여 과거 아카이브와 최근 세션만 남기고 중복을 제거하라.

Frequently Asked Questions about context-compressor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I prevent Lost-in-the-Middle issues when loading large documents into Claude AI sessions?▼

Context compression for Claude AI sessions involves archiving past session logs and slimming recent documents to reduce initial context consumption. This prevents Lost-in-the-Middle by maintaining a single source of truth and keeping startup context under 1,000 lines through a phased workflow.

What is the best way to archive past Claude session logs and reduce startup context size?▼

The best way to archive past Claude session logs is to execute a phased workflow that sequentially measures, slims, compresses, archives, and deduplicates session context. This produces slimmed original documents, archived past records, and a context audit report to prevent Lost-in-the-Middle.

How do I manage CLAUDE.md and WORK_HANDOFF.md files to avoid excessive context consumption at startup?▼

Manage CLAUDE.md and WORK_HANDOFF.md files by applying context compression that deduplicates past archives and recent session logs. This phased workflow slims the original documents and outputs a context audit report, ensuring startup context remains under 1,000 lines.

Can I use a context compression workflow to maintain a single source of truth for Claude session documentation?▼

Yes, you can use a context compression workflow to maintain a single source of truth for Claude session documentation. It enforces a single source of truth by deduplicating past archives and recent sessions, outputting slimmed originals and a context audit report to prevent Lost-in-the-Middle.

What are the limitations of context compression for preventing Lost-in-the-Middle in Claude sessions?▼

The limitation of context compression for preventing Lost-in-the-Middle is that it is specifically applicable to Claude-powered sessions that load large documents at startup. It requires archiving past sessions and slimming the most recent ones to maintain initial context under 1,000 lines.