Deduplicate and synthesize concept stubs into a tiered intellectual map.
npx skills add https://github.com/GYF0311/lorekit --skill concept-synthesis-gyf0311
Please help me install this Agent Skill. Skill: concept-synthesis Source: https://github.com/GYF0311/lorekit/tree/main/brain/skills/concept-synthesis Command: npx skills add https://github.com/GYF0311/lorekit --skill concept-synthesis-gyf0311
It eliminates duplicate, shallow concept pages created during ingestion by deduplicating and synthesizing them into a persistent, tiered intellectual map that traces how ideas evolve over time.
Run concept synthesis on your brain corpus to dedupe concept stubs, assign tiers, and generate a clustered intellectual map under concepts/.
High-intent search queries and answers about installing and using this skill.
Deduplicate concept stubs using Jaccard and substring matching with semantic duplicate detection to merge pages, preserving aliases and timelines. This synthesizes repeated mentions into a persistent knowledge graph tracing idea evolution.
Concept tiering scores frequency, timespan, breadth, and engagement to assign T1–T4 levels, categorizing concepts as Canon, Developing, Speculative, or Riff to structure an intellectual map and filter high-signal ideas for LLM synthesis.
Generate a clustered intellectual map by running LLM synthesis on T1/T2 concepts, then clustering tiered concepts into named domains with summaries and a master concepts/README.md written under the concepts/ directory.
Yes, concept synthesis applies deterministic dedup and merging to large-scale note corpora where repeated or conflicting concept mentions have been ingested as individual stub pages, eliminating shallow duplicates.
LLM synthesis is restricted to T1/T2 high-signal tiers only. T3 and T4 concepts bypass LLM synthesis, relying solely on deterministic dedup, scoring, and tier assignment to preserve processing efficiency.
weipingDeduplicate concept stubs, assign tiers, and generate a tiered intellectual map.
beyonaiDeduplicate concept stubs and cluster them into tiered intellectual maps.
Morris-UTRUSTDeduplicate concept stubs into tiered maps with frontmatter-guided output.
bish-xDeduplicate and tier concept stubs into a structured intellectual map.
