A technical framework and toolkit for building resilient SQLite databases that handle semi-structured, multi-source data through a flexible three-layer schema model.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install flexible-database-design
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install flexible-database-design using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Flexible Database Design is a specialized methodology designed to bridge the gap between rigid relational structures and the unpredictability of semi-structured data. By utilizing Openclaw Skills, this approach implements a Hard Trunk, Soft Tail strategy where core metadata like timestamps and sources are stored in fixed columns, while secondary, varying data is captured in JSON or key-value pairs. This ensures your database remains adaptable to new data types without requiring constant schema migrations.
This skill provides an end-to-end workflow for developers to build sophisticated archives, from personal knowledge bases to complex policy tracking systems. It emphasizes a three-layer data evolution: an original raw layer for traceability, a soft field layer for flexible querying, and a business view layer for high-performance reporting. Using Openclaw Skills allows for a standardized yet highly customizable way to manage information fragmentation across various domains.
To get started with this skill, ensure you have python3 installed on your system. Openclaw Skills provide the necessary scripts and templates for immediate deployment.
# Navigate to your project directory
# Archive your first test record
python3 scripts/archive_item.py -c "Initial test entry" -s "manual"
# Query your stored items
python3 scripts/query_items.py --list
# Perform a field-based search
python3 scripts/query_items.py --field "tags" --value "work"
The architecture follows a strict hierarchy to ensure data integrity and query efficiency:
| Layer | Component | Description |
|---|---|---|
| Raw Layer | raw_content |
Stores the original, unaltered data with hash-based deduplication and source versioning. |
| Soft Field Layer | properties (JSON) |
Holds structured but variable data points (e.g., project IDs, specific report metrics). |
| Business View Layer | Materialized Views | High-performance views that flatten common JSON keys into indexed columns for rapid retrieval. |
Standard trunk fields typically include: id, created_at, source_type, and content_type.
--llm-extract flag during the archiving process.extractors/ directory for specialized file formats.Loading
A technical framework for configuring ESP32-S3-CAM modules as high-performance visual sensors for AI agents.

A specialized skill for safely modifying and managing the configuration of OpenClaw agents and integrations.

A precision debugging skill that captures runtime execution traces to provide deep visibility into application failures for rapid root cause analysis.

A sophisticated runtime debugging tool that enables AI agents to capture, analyze, and resolve software bugs through deep execution tracing and state inspection.

A specialized framework for DTC electronics brands to design high-impact review collection flows and strategic social proof placement to drive conversions.

An automated tool for generating professional, tax-compliant PDF invoices from natural language or structured data commands.








































