How gagan-kikkeri/astra works
ASTRA Sentinel is an OSINT triage and briefing web application. Its main workflow accepts text or a URL, resolves article content, checks for duplicates, classifies and extracts intelligence, synthesizes a briefing, and stores the dispatch for search and review. The README documents additional RSS ingestion, cross-document synthesis, SITREPs, and localization. The graph includes these capabilities, but unsampled implementation details and some route-level wiring are based on documentation rather than inspected source.
Main entry points: main.py REST API · processor.py Dispatch processing · database.py Dispatch database · intelligence.py Briefing synthesis
Components 19
User interface and API
- Agent interface
app/templates/index.html - REST API
app/main.py - Request schemas
app/models.py
Dispatch ingestion
- URL resolution
app/processor.py - Dispatch processing
app/processor.py - Hash deduplication
app/processor.py - Domain triage
app/processor.py - Entity extraction
app/processor.py - RSS feed sync
app/processor.py - LLM gateway
app/llm_gateway.py
Analysis and retrieval
- Briefing synthesis
app/intelligence.py - Cross-document dossiers
app/intelligence.py - Full-text search
app/database.py - Localization
app/translator.py
Storage and configuration
- Dispatch database
app/database.py - Runtime settings
app/config.py
Other components
- Analyst
- Defense RSS feeds
- LLM providers
Connections 25
- Analyst → Agent interface: submits dispatch
README.md - Agent interface → REST API: sends requests
README.md - REST API → Request schemas: validates input
app/main.py - REST API → Dispatch processing: analyzes dispatch
app/main.py - Dispatch processing → URL resolution: resolves URLs
README.md - Dispatch processing → Hash deduplication: checks hash
README.md - Dispatch processing → Domain triage: triages domain
README.md - Dispatch processing → Entity extraction: extracts entities
README.md - Dispatch processing → LLM gateway: requests analysis
app/processor.py - LLM gateway → LLM providers: calls providers
app/llm_gateway.py - Dispatch processing → Dispatch database: inserts dispatch
app/processor.py - REST API → RSS feed sync: syncs feeds
app/main.py - RSS feed sync → Defense RSS feeds: fetches entries
- RSS feed sync → Dispatch database: indexes dispatches
app/processor.py - REST API → Briefing synthesis: generates SITREP
README.md - REST API → Cross-document dossiers: requests synthesis
app/main.py - Cross-document dossiers → Dispatch database: retrieves dispatches
app/intelligence.py - REST API → Full-text search: searches dispatches
README.md - Full-text search → Dispatch database: queries index
app/database.py - REST API → Dispatch database: reads dispatches
app/main.py - Briefing synthesis → Full-text search: retrieves evidence
app/intelligence.py - REST API → Localization: localizes results
app/main.py - Briefing synthesis → Localization: localizes briefing
app/intelligence.py - Dispatch processing → Runtime settings: uses settings
app/processor.py - Dispatch database → Runtime settings: reads database path
app/database.py