Retrieve dynamic memory links across Zettelkasten-style graphs with UHMS.
npx skills add https://github.com/Acosmi/CrabClaw --skill metacog-links
Please help me install this Agent Skill. Skill: metacog-links Source: https://github.com/Acosmi/CrabClaw/tree/main/docs/skills/tools/memory/metacog-links Command: npx skills add https://github.com/Acosmi/CrabClaw --skill metacog-links
Dynamic memory linking across a Zettelkasten-style memory graph surfaces relevant associations, helps maintain context, and accelerates long-term recall in complex knowledge tasks.
Query the memory graph for links related to a given memory ID using memory.metacog.links.get and review the top results.
High-intent search queries and answers about installing and using this skill.
Automatic memory linking in a Zettelkasten graph establishes cross-references for new notes based on similarity and intent signals to surface relevant associations and maintain context.
Retrieving related concepts from a memory graph uses high-dimensional vector search and embedding-based queries to fetch top matches with measured link strength.
BFS-based traversal for exploring contextual pathways in a knowledge graph traverses links up to two hops to reveal connections while applying time-decay to memory strengths.
Filtering memory links by predefined types supports categories including derived_from, contradicts, supports, elaborates, and predicted_by within the memory metacognition system.
Dynamic memory linking features require a UHMS-based memory store to filter results by strength and recency and ensure links remain within defined memory groups.
Memory link strength decays over time in a dynamic knowledge graph because time-decay scoring is applied to memory strengths during BFS traversal to prioritize recent associations.
mrsorbateOptimize AgentDB vector database performance with quantization, HNSW indexing, and batch processing.
SafeAI-Lab-XEnforce data-leakage and permission policies for Feishu interactions.
ever-oliCreate FAISS indexes and run k-nearest-neighbor queries on large embeddings.
Imad-OuteAudits knowledge graph topology for link density, isolated nodes, MOC coverage, and cross-layer gaps.
