guardian-core_skill

Overview

This skill is a minimal, local-first guardian scanner that bundles signature definitions and a lightweight dashboard for on-host threat detection. It offers realtime pre-scan and batch scan/report capabilities with local SQLite logging and no external webhook, API server, or cron automation dependencies. The package is focused on offline, self-contained scanning suitable for privacy-conscious deployments.

How this skill works

Guardian Core runs a realtime pre-scan on incoming content and can perform scheduled or ad-hoc batch scans to generate reports. It uses bundled signature definitions to detect threats, logs events to a local guardian.db SQLite file, and serves static dashboard files for local review. A simple Python API exposes RealtimeGuard for programmatic scanning and decision logic.

When to use it

• You need an on-host scanner with no external network dependencies or cloud integrations.
• You want realtime filtering for chat or messaging inputs before processing.
• You require local auditing of scan results in a lightweight database.
• You prefer a bundled set of signature rules and an easy-to-run dashboard.
• You are deploying in environments with strict data residency or privacy requirements.

Best practices

• Run install.sh under the designated skills directory and verify with provided admin commands.
• Keep the host system patched and limit access to guardian.db to trusted users.
• Use the Python RealtimeGuard API to integrate scans into message handling pipelines.
• Export batch reports regularly and archive them if long-term retention is needed.
• Treat bundled signatures as a baseline; review and extend rules to match your threat model.

Example use cases

• Pre-filter incoming Telegram or chat messages to block or flag abusive content before processing.
• Perform periodic batch scans across archived messages and generate incident reports for review.
• Deploy on an air-gapped system where no webhook or external API calls are allowed.
• Maintain a local dashboard for security ops to inspect recent detections and trends.
• Integrate RealtimeGuard into a bot workflow to decide programmatic blocking or notification actions.