context-tiering

Organizes conversation context into tiered memory with working,Cyclic, reference and archival stores.

15|5|Updated May 17, 2026

Search Tags:#context #llm #agent #retrieval #persistence #tiering #memory-management

Ready to run

npx skills add https://github.com/cropsgg/skills --skill context-tiering

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: context-tiering
Source: https://github.com/cropsgg/skills/tree/main/skills/planning/context-tiering
Command: npx skills add https://github.com/cropsgg/skills --skill context-tiering

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Tiered memory management addresses context-window overflow by organizing information into working, reference, and archival tiers, preserving critical data across long-running sessions and reducing token churn.

Core Features & Use Cases

Tier classification for current, reference, and archival contexts.
Proactive promotion/demotion to optimize memory usage during long sessions.
Efficient retrieval via an indexed reference store and compressed archival storage.
Suitable for long-running agent tasks, multi-file projects, and cross-session continuity.

Quick Start

Promote essential context to the reference tier when the working window approaches capacity, compress and archive infrequently used material, and retrieve from all tiers on demand.

Frequently Asked Questions about context-tiering

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

FAQPage Schema

How do I prevent context window overflow in long-running LLM agent tasks?▼

Prevent context window overflow by organizing long-running conversation context into tiered memory with working, reference, and archival tiers. This approach promotes and demotes elements as the window fills to preserve critical data and reduce token churn.

How does tiered memory management work for multi-file agent sessions?▼

Tiered memory management for multi-file sessions works by classifying context into working, reference, and archival tiers. It enforces proactive promotion and demotion of elements as the window fills, retrieving from indexed tiers on demand.

What is the best way to maintain cross-session continuity for LLM agents?▼

The best way to maintain cross-session continuity is using a tiered memory system that compresses infrequently used material into an archival tier. It balances performance, recoverability, and cost via lossy compression and an indexed reference store.

When do I need to compress and archive infrequently used conversation context?▼

You need to compress and archive infrequently used context when the working window approaches capacity. Proactive tiering moves essential context to the reference tier and applies lossy archival compression to optimize memory usage during long sessions.

Does tiered context memory support multi-turn conversations across multiple modules?▼

Yes, tiered context memory supports multi-turn conversations across multiple files or modules. It applies during multi-turn sessions by enforcing proactive tiering and an indexed reference store to prevent context loss.

Why does my LLM agent lose critical data during long multi-turn sessions?▼

An LLM agent loses critical data during long sessions due to context-window overflow without tiered memory. Organizing information into working, reference, and archival tiers with proactive promotion and demotion preserves data and reduces token churn.