Extract atomic personal facts from conversations and store them in SQLite.
npx skills add https://github.com/arunoda/OkBrain --skill fact-extraction-arunoda
Please help me install this Agent Skill. Skill: Fact Extraction Source: https://github.com/arunoda/OkBrain/tree/main/.agent/skills/fact_extraction Command: npx skills add https://github.com/arunoda/OkBrain --skill fact-extraction-arunoda
This Skill automatically extracts key personal facts from your conversations, stores them, and uses them to personalize future AI interactions, making the AI more aware of your preferences and history.
Ask the AI to extract facts from your conversations to improve its understanding of you.
High-intent search queries and answers about installing and using this skill.
You extract personal facts from chat history by running a background worker process that uses AI models like Gemini or Grok to identify atomic facts. These facts are then injected into subsequent conversations to provide personalized context.
Conversation memory and fact extraction works by analyzing your chat history to identify personal facts, categorizing them into groups like core, technical, project, or transient, and storing them in a SQLite database for future context injection.
Yes, you can use Gemini or Grok to extract user profile facts from conversations. The worker process leverages these AI models to identify atomic personal facts and stores them in a SQLite database for deduplication and context management.
Yes, automatic fact extraction manages deduplication and categorizes extracted facts into core, technical, project, and transient categories. This organization ensures relevant background information is stored efficiently in the SQLite database without redundant entries.
To start extracting facts from your conversations to improve AI context, simply ask the AI to extract facts. A background worker process will then analyze the chat history, identify atomic facts, and store them for future personalization.
Using a SQLite database for conversation memory limits fact storage to local single-file capacity and restricts concurrent read/write operations during the background extraction process. Fact categorization and deduplication are managed within this local database structure.
majesticlabs-devCompute AI visibility metrics for brand recommendations across AI chat platforms.
arunodaSearch extracted facts locally using Ollama embeddings and sqlite-vec.
KoulbSearch arXiv and Google Scholar for academic papers and store results in a local SQLite database.
thecloudtipsQuery a local SQLite memory store and inject relevant past interactions into plugin execution context.
