knowledge-base

What This Skill Does

The Knowledge Base skill serves as the central nervous system for your personal data, utilizing SQLite with FTS5 (Full-Text Search) to index, organize, and retrieve information across fragmented sources. By centralizing WhatsApp contact data, local markdown-based documentation, and exported ChatGPT conversation history, it transforms static files into a searchable, persistent memory system. Whether you are performing complex contact lookups, digging through historical chat logs, or referencing technical documentation, this skill provides instant, local access to your digital footprint.

Installation

To install this skill, use the ClawHub CLI command: clawhub install openclaw/skills/skills/globalcaos/knowledge-base

Once installed, you must initialize your database. Navigate to the skill directory and run: python3 skills/knowledge-base/scripts/init_db.py

This will create a db/jarvis.db file in your workspace root. You can import VCF contact files using import_vcf.py or place your ChatGPT exports and markdown files in their respective folders to populate the search index.

Use Cases

• Unified Contact Management: Quickly identify which WhatsApp groups a specific contact belongs to, or find a person by their phone number across disparate contact lists.
• Long-term Conversation Recall: Search deep into your past ChatGPT sessions for specific technical insights or brainstormed ideas that are otherwise buried in export files.
• Personal Documentation Wiki: Keep all your research and notes in simple markdown files within the memory/ directory and query them naturally using the search command.
• Data Analytics: Perform ad-hoc SQL analysis on your communication metadata to understand your connectivity patterns or group memberships.

Example Prompts

1. "Search my knowledge base for any conversations regarding my project 'Titan' and show me the relevant ChatGPT messages."
2. "Look up the contact info for +34659418105 and list all the WhatsApp groups we share."
3. "Query my documents for the technical notes I wrote last month about database schema design."

Tips & Limitations

• Indexing: This skill is local-first. If you add new markdown documents or new export files, you must run the sync script or re-initialize the database to ensure the search index reflects the updates.
• Performance: While SQLite FTS5 is highly efficient, ensure your chatgpt-export directory is clean; extremely large JSON files may increase initialization time.
• Formatting: Keep your markdown files in memory/ well-structured to improve the relevance of search results returned by the doc command.