sam-context-size-management

name	SAM: Context Size Management
description	Define explicit guidance for measuring and managing context size per agent. What's the target token budget? How to detect context pressure?
metadata	{"topic":"sam-context-size-management","source":"Gap analysis of SAM framework","added":"2026-02-01","priority":"P2","type":"Feature","status":"open"}

Suggested location: stateless-software-engineering-framework.md (section 2.1 or Appendix C)

Research first: How do agent frameworks measure context usage? What token counting approaches exist? How does Claude Code handle context limits internally?