Collected sources and patterns will appear here. Add from search or the patterns library.
ProjectCodebase + MemoryScaffold → DriftReport → UpdatedScaffold → AgentContext
Maintains a structured file (e.g. AGENTS.md or equivalent) that records current project context, architecture decisions, and relevant file paths for AI coding agents; a CLI detects semantic drift between the memory file and actual codebase state (changed APIs, moved files, renamed functions) and prompts a human to update the scaffold before the agent acts on stale context.
Problem it solves
AI coding agents repeatedly re-derive project context from scratch each session, wasting tokens and making mistakes on stale assumptions; a persistent structured memory file with drift detection keeps context synchronized with the live codebase.
Consumes
Emits
Establishes
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.