How siddardha10/cyber_assistant works
CyberSentinel is a local Windows monitoring assistant. Its main workflow collects system and security signals into SQLite, then answers user questions through offline matching or LLM-assisted query planning; users can interact through a CLI or API. A separate command path translates natural-language requests into validated actions before execution. The README also describes alerts and optional network monitoring and LLM support. Edges reflect the provided excerpts, source-index references, or README; unsampled connections are omitted where their callers are not evidenced.
Main entry points: database.py SQLite access · llm.py LLM query planning · translator.py Command translation · server.py Web API
Components 20
User Interfaces
- App startup
main.py - CLI chat
cli/repl.py - Web API
api/server.py
Monitoring and Alerts
- System monitors
monitors - Alert notifications
alerts/notifier.py
Persistence
- SQLite access
db/database.py - Database schema
db/schema.py
Question Answering
- Query entry point
query/engine.py - Offline matching
query/offline.py - LLM query planning
query/llm.py - SQL validation
query/sql_guard.py - Web fallback
query/web_fallback.py - Query catalog
query/catalog.py
Command Execution
- Command translation
commands/translator.py - Intent routing
commands/intent_router.py - Entity resolution
commands/entity_resolver.py - Path resolution
commands/path_resolver.py - Command templates
commands/templates.py - Command execution
commands/executor.py
Other components
- Local user
Connections 29
- Local user → CLI chat: asks questions
README.md - Local user → Web API: asks questions
README.md - App startup → System monitors: starts monitors
main.py - App startup → CLI chat: starts CLI
main.py - App startup → Web API: starts API
main.py - System monitors → SQLite access: stores observations
monitors/logins.py - System monitors → Alert notifications: raises alerts
- Alert notifications → SQLite access: records alerts
alerts/notifier.py - SQLite access → Database schema: uses schema
db/database.py - CLI chat → Query entry point: asks questions
cli/repl.py - Web API → Query entry point: asks questions
api/server.py - Query entry point → Offline matching: tries matching
query/engine.py - Query entry point → LLM query planning: routes fallback
query/engine.py - Query entry point → SQLite access: logs answers
query/engine.py - LLM query planning → SQLite access: reads query data
query/llm.py - LLM query planning → SQL validation: validates SQL
query/llm.py - LLM query planning → Command translation: translates commands
query/llm.py - LLM query planning → Command execution: executes commands
query/llm.py - LLM query planning → Web fallback: uses fallback
query/llm.py - Web API → Query catalog: serves suggestions
api/server.py - Web API → Command execution: executes approved command
api/server.py - Command translation → Intent routing: routes intent
commands/translator.py - Command translation → Entity resolution: resolves entities
commands/translator.py - Command translation → Path resolution: resolves paths
commands/translator.py - Command translation → Command templates: selects template
commands/translator.py - Command translation → SQLite access: stores corrections
commands/translator.py - Intent routing → Command templates: retrieves actions
commands/intent_router.py - Command execution → SQLite access: logs execution
commands/executor.py - Command execution → Command templates: loads template
commands/executor.py