# Context window

The finite sequence of tokens a model can directly condition on during a single inference operation.

technical-term · Context & knowledge · foundational · Reviewed 2026-08-09

## Definition

The context window contains the active instructions, user input, selected history, retrieved knowledge, tool results, and other state presented to the model. A larger window increases capacity but does not ensure relevance, consistency, truth, or correct attention.

Finite model context is established; effective use remains workload- and model-dependent.

## Why it matters

Context is a scarce runtime resource even when nominal token limits are large.

Selection, ordering, provenance, compression, and removal are engineering decisions.

## System anatomy

- **Capacity:** The maximum token budget available to the inference.
- **Composition:** The mix and ordering of instructions, state, examples, evidence, and retrieved data.
- **Provenance:** Where each context item came from and what trust level it carries.



## Important distinctions

- **Memory:** Memory persists or retrieves information across time; context is what is visible for this decision.
- **Knowledge:** Presence in context does not make a statement true or authoritative.



## Implementation signals

- Load the smallest sufficient context for the decision
- Label source, recency, and trust level
- Measure performance under realistic context length and noise

## Failure modes

- Context stuffing hides the governing requirement
- Stale summaries silently replace primary evidence



## Related knowledge

- [Context engineering](https://aisdlc.ai/agentic-engineering/context-engineering) — The deliberate selection and maintenance of instructions, knowledge, tools, state, examples, and artifacts within a finite model attention budget.
- [Working state](https://aisdlc.ai/agentic-engineering/working-state) — The current, inspectable record of an agent’s task position, observations, decisions, pending actions, and budgets.
- [Retrieval-augmented generation](https://aisdlc.ai/agentic-engineering/retrieval-augmented-generation) — A pattern that retrieves external information at inference time and supplies selected results to a generative model.

## Sources and further study

- [Anthropic — Effective context engineering for AI agents](https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents) — A working model for treating model-visible context as a finite resource that must be selected and maintained. **Use:** First-party case study. **Limitation:** This first-party account documents one organization, product, or implementation context and should not be generalized without local evidence.
- [Peter Yang — 5 Rules for Building AI Agents That Work in Production | Nan Yu & Jacob Shumway](https://www.youtube.com/watch?v=4mKtJzfGj0U) — A Linear engineering discussion of production agents, tool design, context loading, feedback, evaluation, and simple model-in-a-loop explanations. **Use:** Practitioner perspective. **Limitation:** The five rules and “LLM in a loop plus tools” shorthand reflect one product team’s experience; they are not a standard or a complete enterprise-agent definition.

---

This library synthesizes cited research, standards, official documentation, and clearly attributed practitioner perspectives. Maturity describes the state of a concept—not vendor endorsement, production readiness, or permission to deploy. Benchmarks and demonstrations do not replace use-case evaluation, governed controls, independent verification, or named human release authority.