Turn any folder of files, code, docs, images, and videos into a navigable, clustered knowledge graph with a comprehensive audit trail.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install sdw-kb
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install sdw-kb using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is sdw-kb: Knowledge Graph Generation & Community Detection?
The sdw-kb skill is a powerful framework built around a raw folder workflow that transforms unorganized data—including source code, markdown documentation, research papers, images, and media files—into structured, highly discoverable knowledge graphs. By utilizing community detection and GraphRAG-ready outputs, it uncovers hidden, cross-document connections that standard AI tools often fail to capture across distinct sessions.
Operating completely within a self-contained environment via the uv tool manager, this skill ensures total process reproducibility. Every generated node and relationship is preserved persistently, guaranteeing that your knowledge base survives indefinitely and remains queryable without the need to continuously re-read original source documents.
sdw-kb: Knowledge Graph Generation & Community Detection Use Cases
- Architectural onboarding for complex codebases to rapidly visualize system dependencies and hidden pattern connections before drafting modifications.
- Synthesis of vast reading collections containing mixed assets like PDFs, research papers, notes, and tweets into a centralized, searchable concept graph.
- Comprehensive document auditing to visually trace design rationale, engineering tradeoffs, and conceptual provenance using an explicit verification standard.
- Continuous knowledge accretion within a personal raw data bucket, transforming scattered daily thoughts, snapshots, and logs into a clustered network.
How sdw-kb: Knowledge Graph Generation & Community Detection Works
- Path Resolution: Resolves the dedicated working folder path inside the isolated central base directory, supporting explicit naming or folder derivation.
- File Discovery: Performs structural scanning across code, textual documentation, images, and video assets while validating limits and identifying skipped resources.
- Audio Extraction: Automatically processes multi-format audio and video recordings through Whisper transcription engines when media structures are present.
- Parallel Analysis: Triggers structural abstract syntax tree generation in tandem with multi-agent semantic LLM processing for unmapped contextual layers.
- Graph Aggregation: Merges programmatic architecture components with semantic discoveries, building a unified network map while stripping redundant elements.
- Community Partitioning: Executes cluster detection algorithms to group relevant segments together, measuring community cohesion via network metrics.
- Visualization Export: Compiles final relational elements into interactive browser-ready HTML files, raw JSON, and plain-text markdown summaries.
sdw-kb: Knowledge Graph Generation & Community Detection Setup
Ensure you have the uv package manager ready on your system. The Openclaw Skills runner will handle the remaining process automatically.
# Verify or manually trigger graphifyy setup inside the uv tool virtual environment
uv tool install graphifyy
# Run the complete knowledge graph pipeline on your current active folder
/sdw-kb
sdw-kb: Knowledge Graph Generation & Community Detection Data Schema & Taxonomy
All analytical outputs and cached states are written directly to the self-contained knowledge base folder located at ~/.sdw/knowledge_bases/<kb-name>/, maintaining absolute decoupling from your project files.
Output Files Matrix
| Artifact Target | File Path | Objective Purpose |
|---|---|---|
| Interactive UI | graphify-out/graph.html |
Network visualization wrapper for browser exploration. |
| Structured Data | graphify-out/graph.json |
Comprehensive GraphRAG-compliant entity and relationship schema. |
| Audit Summary | graphify-out/GRAPH_REPORT.md |
Plain-language documentation, god nodes, and surprise links. |
| Cost Registry | graphify-out/cost.json |
Token consumption tracking logs across computational cycles. |
Relationship Extraction Taxonomy
- EXTRACTED: Explicit structural relationships such as codebase imports, file citations, and direct function calls (Confidence: 1.0).
- INFERRED: Latent conceptual pairings, implicit shared data structures, or architecturally coupled modules (Confidence: 0.6 - 0.95).
- AMBIGUOUS: Uncertain structural pairings or vague text references flagged for developer peer review (Confidence: 0.1 - 0.3).
sdw-kb: Knowledge Graph Generation & Community Detection Advanced Features
- Parallel Multi-Agent Chunking: Dispatches concurrent background extraction workers to speed up intensive processing operations on large non-code datasets.
- Incremental Re-Extraction: Scans via change manifest checks to exclusively parse modified or newly added assets, keeping cost overheads low.
- Pluggable Visual Adapters: Generates Gephi-compliant GraphML charts, standalone SVG vector maps, and Cypher generation matrices for Neo4j database instances.
- Obsidian Vault Mapping: Automatically transforms structural sub-elements into a markdown vault complete with native interactive canvas representations.
- Integrated Tool Extensions: Offers structural hooks for git workflows, direct Model Context Protocol server exposure, and native CLAUDE.md profile integration.
SKILL.md
Loading
Related Openclaw Skills
Test-Deployment Template Starter Pack
A battle-tested Python starter template to build and deploy production-ready vertical AI agents for Telegram in 10-15 minutes.
Gaokao Volunteer Assistant
An AI-powered gaokao volunteer filling helper using ranking-based matching and risk classification to automate Chinese college selection.
Musk Mindset
A cognitive engineering framework that transforms your AI agent into a rigorous, first-principles-driven thinker utilizing the famous five-step algorithm and business heuristics of Elon Musk.
Business Plan Writer
An advanced AI-powered assistant designed to write comprehensive, investor-ready business plans based on nine core industry-standard modules.
Alibaba Cloud DNS Resolution Diagnostics
An automated diagnostic utility designed to troubleshoot DNS resolution anomalies, domain unreachable issues, and routing misconfigurations across Alibaba Cloud and third-party DNS platforms.
DeepRead Tax Forms
An AI-powered document extraction skill to convert W-2s, 1099s, 1040s, and scanned tax forms into clean, typed JSON with per-field confidence flags.
Featured*
