Memory

To persist user preferences and context across conversational sessions, this Skill stores facts in a structured JSON file. It categorizes memories into identity, preference, project, and environment, ensuring details remain available for future interactions without repetition.

Relevance-based recall retrieves memories by calculating token overlap scoring against the current prompt. This mechanism identifies semantically related facts within the JSON storage, returning the most applicable user context dynamically.

When updating a previously stored fact, the intelligent superseding mechanism automatically replaces the older entry. Newer information on the same subject overwrites past memories to prevent conflicts and maintain accurate context.

Yes, you can track access frequency and timestamps for memory management. The system logs when memories are retrieved and how often they are accessed, providing data to manage and prioritize stored context effectively.

No external dependencies are required to store facts. The skill operates independently using a structured JSON file format to manage and persist user context, preferences, and project details locally.

The best way to categorize project environment details is by using the built-in classification system. The skill automatically organizes context into identity, preference, project, and environment categories to optimize later retrieval.