Web Knowledge Enricher
How HelixMind automatically fetches relevant documentation and web knowledge while you work.
Web Knowledge Enricher
HelixMind automatically searches the internet for relevant documentation while you work. no manual searching needed.
How It Works
| Step | What Happens |
|---|---|
| 1. Topic Detection | Analyzes your message for technology, pattern, or error topics |
| 2. Web Search | Searches DuckDuckGo for relevant results |
| 3. Knowledge Extraction | Extracts code examples, key points, and summaries from pages |
| 4. Spiral Storage | Stored as L6 Web Knowledge nodes with quality scores + embeddings |
| 5. Brain Update | Popup notification + particle burst in 3D Brain visualization |
CLI Notification
🌐 Web Knowledge: JWT best practices for production
Source: https://auth0.com/docs/tokens
Quality: 0.94 │ Stored in L6 (Web Knowledge)
Explicit Web Research
Besides automatic detection, you can explicitly trigger research:
> Search the web for Next.js 15 server actions migration guide
The AI uses the web_research tool to fetch, extract, and store findings.
What Gets Stored
| Field | Description |
|---|---|
content | Extracted knowledge (code, key points, summary) |
source | Original URL |
quality | Quality score (0.0 – 1.0) |
topic | Detected topic that triggered the search |
embedding | 384d vector for semantic search |
Privacy
| Aspect | Detail |
|---|---|
| Search engine | DuckDuckGo (no tracking) |
| Processing | All content processed locally |
| Storage | Only extracted knowledge stored (not raw HTML) |
| External data | Only the search query is sent externally |
| Offline | Automatically disabled with Ollama |