Unverified50% confidenceFactExact time
Smart Connections chunks the notes in the vault, computes and caches embeddings, converts the user's question into a vector when they ask, and uses cosine similarity to find the most relevant fragments to feed to the LLM.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedIn this system, each Slack question becomes a Langfuse trace, recording the question, answer, user, thread session, as well as the prompt, response, and token count for each round of Gemini inference.72% similarUnverified程序通过余弦相似度计算相关段落，动态拼装发送给 AI 的 prompt，实现标准 RAG 流程69% similarUnverifiedA local model (run via Ollama, LM Studio, etc.) keeps the weight files and inference computation on the local machine, and the note text is read from disk and then enters local memory and VRAM, theoretically without going over the network; a remote model is usually called via API, and the retrieved note fragments are sent to the provider's server as request content.68% similarUnverifiedEntering /generate in Cursor Chat will scan the existing codebase, identify patterns and libraries already in use, and automatically generate matching rules files.67% similarUnverifiedMany agent frameworks write the full reasoning chain—including raw tool-call responses—to run logs for debugging purposes.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1016830API
curl https://kongchang.com/api/v1/knowledge/claims/1016830MCP
get_claim(id=1016830)