Definition

What is Context Offloading?

Last updated

The practice of keeping an AI agent's working context window small by moving state to an external destination, such as a file, a sub-agent, or a retrievable store, and bringing it back only when a step needs it.

Long-running agents accumulate conversation history, tool outputs, and observations that dilute attention and degrade accuracy. Context offloading patterns, including structured note-taking, sub-agent delegation, and just-in-time retrieval, relocate that bulk to a destination so the agent works with a small, high-signal window and can recover detail on demand. It differs from reduction techniques like compaction, which shrink the window in place rather than relocating anything.

Related Context Window Context Compression Context Rot Context Engineering Multi-Agent System

Further reading

Articles about Context Offloading

Context bloat: why long-running agents break

Context bloat is when accumulated tool-call output crowds out an agent's task. Tool calls, not window size, break long-running agents. Here is the fix.

Demand paging for the AI context window

A 2026 systems paper found 21.8% of tokens in agent context windows are wasted. Demand paging treats the AI context window as L1 cache, not full memory.

MCP Tasks: long-running work as context offloading

MCP Tasks let a server return a durable handle instead of a blocking result, keeping a long-running tool call's interim state off the agent's context window.

AI notetakers ship the wrong artifact

AI notetakers ship transcripts, but downstream work needs decisions, drafts, or handoffs. The artifact gap is a context engineering problem, not transcription.

View all 5 articles →

All terms

View full glossary
Agent Drift Agent Reliability AI Agent AI Hallucination AI Second Brain Chunking Context as a Service Context Bloat Context Compression Context Container Context Drift Context Engineering Context Offloading Context Poisoning Context Portability Context Pruning Context Rot Context Window Epistemic Provenance Fine-Tuning Knowledge Graph MCP Resources MCP Server MCP (Model Context Protocol) Multi-Agent System Prompt Caching Prompt Engineering RAG (Retrieval-Augmented Generation) Semantic Search Structured Context Tool Poisoning Wire

Put context into practice

Create your first context container and connect it to your AI tools in minutes.

Create Your First Container