How tintunaung124/opennews-auditor works
OpenNews Auditor takes an article URL through a browser interface, extracts its content and claims, gathers external reporting, verifies claims with AI, and presents an explainable Trust Score. The backend’s domain stages are described in the README but implemented in the single sampled backend module; the graph separates them for clarity. The excerpts show the browser request and result handling, but not the backend audit route or its internal calls, so those backend relationships rely on README documentation.
Main entry points: main.py Audit results · script.js Request and rendering · index.html Audit interface
Components 15
Browser client
- Audit interface
frontend/index.html - Request and rendering
frontend/script.js
Audit workflow
- Audit endpoint
backend/main.py - Article retrieval
backend/main.py - Content extraction
backend/main.py - Claim extraction
backend/main.py
Evidence and verification
- News search
backend/main.py - Evidence preparation
backend/main.py - Claim verification
backend/main.py
Scoring and results
- Trust scoring
backend/main.py - Score explanation
backend/main.py - Audit results
backend/main.py
Other components
- Article submitter
- External news
- AI providers
Connections 17
- Article submitter → Audit interface: opens
README.md - Audit interface → Request and rendering: uses
frontend/script.js - Request and rendering → Audit endpoint: posts URL
frontend/script.js - Audit endpoint → Article retrieval: retrieves
README.md - Article retrieval → Content extraction: extracts content
README.md - Content extraction → Claim extraction: identifies claims
README.md - Claim extraction → News search: searches claims
README.md - News search → External news: searches
README.md - News search → Evidence preparation: gathers evidence
README.md - Evidence preparation → Claim verification: provides evidence
README.md - Claim verification → AI providers: verifies claims
README.md - Claim verification → Trust scoring: supplies results
README.md - Trust scoring → Score explanation: explains score
README.md - Trust scoring → Audit results: includes score
README.md - Score explanation → Audit results: includes explanation
README.md - Audit endpoint → Audit results: returns results
backend/main.py - Request and rendering → Audit results: renders response
frontend/script.js