personal-kb-updater

Archives conversation insights into a structured personal knowledge base with tiered retrieval.

2|Updated Aug 5, 2026
One-click install
npx skills add https://github.com/YuiClein666/digital-twin-kit --skill personal-kb-updater-yuiclein666
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: personal-kb-updater
Source: https://github.com/YuiClein666/digital-twin-kit/tree/main/skill
Command: npx skills add https://github.com/YuiClein666/digital-twin-kit --skill personal-kb-updater-yuiclein666

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI assistants forget everything between sessions, forcing you to reintroduce yourself repeatedly. This Skill gives the AI a persistent, structured memory of you—your experiences, personality, values, decisions, and recent life events—stored in a local Markdown knowledge base that you fully own. ## Core Features & Use Cases - Tiered Retrieval Strategy: A four-level (L0–L3) decision tree determines when and how deeply to read the knowledge base before each reply, balancing context depth against cost. - Three Update Triggers: Manual commands ("记一下"), automatic summaries every 10 conversation turns, and proactive AI observations of your behavioral patterns written to an observations file. - Dedup & Contradiction Handling: Merges repeated facts, upgrades confidence levels, and preserves contradictory statements as evolution history rather than deleting old judgments. - Cognitive Companionship: When you feel stuck, it replays your past decisions and stated values to offer grounded options instead of generic advice. - Use Case: After weeks of chatting, you ask "Should I take this job offer?" The AI reads your values file and past major decisions, then shows how this choice aligns or conflicts with what you said mattered most. ## Quick Start Say "记一下:我今天决定接受新公司的 offer" and the AI will archive the decision into the knowledge base with confidence tagging and follow-up questions.

Frequently Asked Questions about personal-kb-updater

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make an AI remember things across conversations?

Use this Skill with a local knowledge-base directory of Markdown files. The AI archives facts, decisions, and observations into categorized files, then reads them back before replying, so memory persists across sessions and even across different AI tools.

How does the tiered knowledge base retrieval work?

A decision tree classifies each message into L0 (no lookup), L1 (scan index tags), L2 (read relevant category files), or L3 (read core files plus topic files). Personal keywords, triggers, and every tenth turn escalate the retrieval level automatically.

Can I use this personal knowledge base with different AI tools?

Yes. The rules live in a portable AI-SKILL.md file with no dependency on any vendor memory system. Clone the knowledge-base folder, paste the skill instructions into any AI tool, and point it at the same directory path.

What happens when new information contradicts old records?

Old records are never deleted. Contradictions are appended with timestamps and marked as revised, preserving the evolution of judgments over time. Confidence levels upgrade when the user confirms earlier low-confidence observations.

Where is sensitive personal data stored in the knowledge base?

Sensitive items like phone numbers, ID numbers, and addresses go into a private/ subdirectory that is excluded from git via .gitignore. The main knowledge base files stay in your own local folder or private repository only.