Luca Sacchi Ricciardi 5aab19626f feat: implement OpenRouter AI integration in n8n workflow ...

Add 'Call OpenRouter' node to LogWhisperer_Ingest workflow:

New Node Features:
- Model: openai/gpt-4o-mini via OpenRouter API
- System prompt with Metodo Sacchi (Safety First, Little Often, Double Check)
- Timeout: 10 seconds with AbortController
- Log truncation: max 2000 characters
- Required headers: Authorization, HTTP-Referer, X-Title
- Error handling with graceful fallback response
- Output: JSON with ai_analysis, ai_status, ai_timestamp, ai_model

Workflow Flow:
Webhook → HMAC Validation → Data Validation → Store Log → Call OpenRouter → Critical Severity Check → Response

Test Suite (workflows/test_openrouter.js):
- 10 comprehensive tests covering:
  - Input/output structure validation
  - Log truncation logic
  - OpenRouter API payload format
  - Required HTTP headers
  - AI response structure
  - Fallback error handling
  - Timeout configuration
  - Dangerous command patterns
  - System Prompt Metodo Sacchi validation
  - Workflow connections

Environment Variables Required:
- OPENROUTER_API_KEY
- OPENROUTER_SITE_URL (optional, defaults to https://logwhisperer.ai)
- OPENROUTER_APP_NAME (optional, defaults to LogWhispererAI)

Next Steps:
1. Configure environment variables in n8n
2. Import updated workflow to n8n instance
3. Configure PostgreSQL credentials
4. Test with sample log payload

Refs: docs/specs/ai_pipeline.md (section 4.1)

2026-04-02 19:40:42 +02:00

13 KiB

Raw Blame History

View Raw