skills.rest
guanyang
guanyangCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

context-compression

Compress conversational data in AI agent sessions to optimize token usage.

947|163|Updated Jan 13, 2026
Search Tags:#summarization#token optimization#agent memory#context compression#llm efficiency#long context
One-click install
npx skills add https://github.com/guanyang/antigravity-skills --skill context-compression-guanyang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: context-compression
Source: https://github.com/guanyang/antigravity-skills/tree/main/skills/context-compression
Command: npx skills add https://github.com/guanyang/antigravity-skills --skill context-compression-guanyang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of managing large amounts of conversational data in AI agent sessions, preventing context window overflows and ensuring efficient token usage.

Core Features & Use Cases

  • Context Compression: Implements strategies to reduce the token count of conversation history.
  • Token Optimization: Focuses on minimizing tokens per task, not just per request, to improve overall efficiency.
  • Use Case: In a long-running AI coding session that generates millions of tokens, this Skill can compress the history to keep the agent focused and within operational limits, preventing it from "forgetting" critical details.

Quick Start

Use the context-compression skill to summarize the conversation history to reduce token usage.

Frequently Asked Questions about context-compression

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

FAQPage Schema
How do I optimize token usage and prevent context window overflow in long AI agent sessions?▼

To prevent context window overflow, you can compress large volumes of conversational data using anchored iterative summarization and opaque compression. This optimizes token usage per task and keeps the agent focused within operational limits.

What is anchored iterative summarization for context compression?▼

Anchored iterative summarization is a context compression technique that reduces conversation history token count. It balances token savings against information loss by regenerating full summaries, preventing coding agents from forgetting critical details.

How do I stop my coding agent from forgetting details after generating millions of tokens?▼

Stop agent memory loss by compressing long conversation histories with regenerative full summaries. This addresses the artifact trail problem, ensuring critical coding details are retained without exceeding token limits.

Does context compression work for long-running AI coding sessions?▼

Yes, context compression is designed for long-running AI coding sessions. It minimizes tokens per task rather than just per request, effectively managing millions of generated tokens to maintain agent operational efficiency.

What are the limitations of using opaque compression for agent memory?▼

The main limitation of opaque compression for agent memory is the potential for information loss. It requires balancing token savings against context retention, utilizing regenerative full summaries to mitigate critical detail forgetting.

Related Skills

Explore More Software Engineering Skills→
gaoqiongxiegaoqiongxie
1

claude-mem

Capture session key points and search historical project context.

Community
Advanced
TwuanMinnTwuanMinn

context-window-management

Summarize, trim, route, and prioritize content to manage LLM context windows.

Community
Intermediate
HemantSudarshanHemantSudarshan

context-window-management

Optimizes LLM context windows by summarizing, prioritizing, and selectively trimming content.

Community
Advanced
marinvchmarinvch

context-compression

Compress long agent conversations into anchored structured summaries preserving session context.

Community
Advanced