context-memory

To keep AI context persistent across sessions, use persistent memory files to organize user preferences, project background, workflows, errors, and learnings, ensuring reliable context recovery at task startup without relying on temporary conversation memory.

To prevent forgetting user preferences when switching topics, implement persistent memory files that store long-term context and tool settings, allowing the AI to review relevant memory and apply established preferences before executing new tasks.

The workflow for recovering project background involves a structured startup process: classify the request, determine if it is a new task, review relevant memory and operational guidance, execute the work, and record important updates or corrections in the appropriate memory files.

Persistent AI memory works effectively for long-running projects and frequent topic switching by organizing daily logs and learnings into maintainable files, allowing the assistant to recover relevant context and build continuously on prior work.

To record AI mistakes and avoid repeating documented errors, capture user corrections, mistakes, and important decisions in continuous learning memory files during execution, ensuring future tasks can reference and build upon prior work safely.

You should not use persistent memory files for AI collaboration context if your work involves isolated, single-session tasks without recurring topics, as the structured startup process for task classification and memory review adds overhead unnecessary for ephemeral workflows.