Store information into working, episodic, or semantic memory tiers.
npx skills add https://github.com/X-McKay/kubani --skill store-memory
Please help me install this Agent Skill. Skill: store-memory Source: https://github.com/X-McKay/kubani/tree/main/kubani/skills/general/memory/store-memory Command: npx skills add https://github.com/X-McKay/kubani --skill store-memory
This Skill addresses the challenge of efficiently storing and retrieving information by providing a structured way to manage data across different memory tiers.
Use the store-memory skill to save the observation that the 'api-gateway' pod is experiencing high latency.
High-intent search queries and answers about installing and using this skill.
Hierarchical memory storage organizes information into working, episodic, and semantic tiers. It classifies and persists extracted facts with appropriate metadata, enabling efficient retrieval and pattern identification.
You store observations by saving event details in episodic memory. The system extracts structured facts from raw content using an LLM and persists the data with metadata for later analysis or pattern identification.
Structured fact extraction leverages an LLM to parse raw content and identify key facts. These extracted facts are then persisted into the memory system with appropriate metadata, making the data more usable for retrieval.
Yes, you can identify permanent patterns by storing recent events in episodic memory. The system supports learning from these observations and classifying learned permanent patterns into the semantic memory tier.
The best way to organize ephemeral versus permanent data is through a hierarchical memory system. It automatically classifies ephemeral data into working memory and permanent patterns into semantic memory, ensuring structured management.
You should avoid a hierarchical memory system when your data does not require tiered classification or pattern learning. If your use case only involves simple, flat storage without needing episodic or semantic distinction, this approach is unnecessary.
bettercallfanMap LLM visual observations into video monitoring Event Candidate semantics.
X-McKaySearch hierarchical memory tiers and return relevance-ranked results.
just-cameronObserve ATProtocol firehose activity and record engagement patterns as cognition thoughts.
neuron-coreIntegrate vector stores, embeddings, and document loaders into Neuron AI agents.
