skills.rest
mikeparcewski
mikeparcewskiCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

context-engineering

Optimize context windows, memory scopes, and token usage in multi-agent systems.

8|2|Updated Feb 9, 2026
Search Tags:#context#multi-agent#strategy#token-optimization#prompt-engineering#memory-management
One-click install
npx skills add https://github.com/mikeparcewski/wicked-garden --skill context-engineering-mikeparcewski
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: context-engineering
Source: https://github.com/mikeparcewski/wicked-garden/tree/main/skills/agentic/context-engineering
Command: npx skills add https://github.com/mikeparcewski/wicked-garden --skill context-engineering-mikeparcewski

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Context window management, token optimization, and memory design for efficient multi-agent systems. The goal is to maintain coherent, relevant context while controlling token usage across tasks, sessions, and failed runs.

Core Features & Use Cases

  • Context window sizing across models and tasks (anticipatory, Just-in-Time, hybrid) to minimize token usage without sacrificing task fidelity.
  • Memory pattern guidance for short-term working context, long-term storage with vector representations, and episodic summaries to preserve decisions and outcomes.
  • Prompting and loading strategies, including selective loading, caching, and progressive detail loading, to optimize retrieval and execution cost.

Quick Start

Configure a context-loading plan for a multi-agent workflow that minimizes token usage while preserving task fidelity.

Frequently Asked Questions about context-engineering

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

FAQPage Schema
How do I optimize token usage across multi-agent systems without losing context?▼

Optimize token usage by applying context window sizing across anticipatory and just-in-time loading strategies, categorizing memory into short-term, long-term, and episodic scopes to maintain task fidelity while controlling costs.

What is the best way to manage memory scopes for multi-agent pipelines?▼

Categorize multi-agent memory into short-term working context, long-term storage with vector representations, and episodic summaries to preserve decisions and outcomes across agent pipelines.

When should I use just-in-time context loading instead of anticipatory loading?▼

Use just-in-time context loading to minimize token usage by retrieving information only when needed during execution, whereas anticipatory loading pre-fetches context to ensure immediate availability across agent tasks.

Can I apply progressive detail loading to reduce prompt execution costs?▼

Apply progressive detail loading to retrieve context incrementally as task complexity increases, reducing prompt execution costs by avoiding unnecessary token consumption during multi-agent workflows.

How does caching improve context window management for multi-agent workflows?▼

Caching improves context window management by storing frequently accessed context, reducing redundant retrieval operations and lowering token usage across repeated multi-agent pipeline executions.

Related Skills

Explore More Software Engineering Skills→
haidonglethqbhaidonglethqb

context-engineering

Optimize LLM context windows with deterministic token budgeting and probe-based evaluation.

Community
Advanced
Shakudo-ioShakudo-io

context-engineering-collection

Identify, categorize, and deploy Agent Skills for context engineering in AI agents.

Official
Advanced
jokken79jokken79
1

agent-memory-systems

Coordinate short-term context, long-term memory, and retrieval strategies for AI agents.

Community
Advanced
CAPHTECHCAPHTECH
1

pce-compact

Compress extended session histories into summarized notes for ongoing projects.

Official
Intermediate