INFILTRATOR is a hardened Tkinter control panel for recursive acquisition and passive target analysis. Each HTTP request is a fresh identity: new user-agent, new connection, Tor SOCKS routing. Transparent. Auditable. No silent behavior.
Built for controlled, explicit execution. No hidden services. No background polling. If it moves, it moves because you told it to.
Fresh TCP connection and randomized User-Agent for every request. No keep-alive. No pooling. No sticky identity.
Routes traffic through 127.0.0.1:9050 using SOCKS5h (DNS through Tor).
Preserves remote paths locally under SCRAPES/<domain>/… for clean offline inspection.
Passive, read-only interrogation of selected targets. Extracts title, meta description, and OpenGraph tags. Prints everything to stdout, mirrored in the UI.
ANALYZE SELECTED (READ ONLY) - tor-routed fetch - header + html retrieval - stdlib HTML parsing - no DB mutation - no silent writes
GUI-first control panel. Add targets to DaRK.db, select, run. STOP is immediate and kills process trees cleanly.
BASIC FLOW 1) Add URL → stored in DaRK.db 2) Select one or more targets 3) Scrape Selected / Scrape ALL 4) (Optional) Analyze Selected 5) STOP aborts the run immediately
Identity changes happen at request boundaries, not “sessions.” This makes tracking scripts see what they fear: no stable fingerprint to hold onto.
REQUEST BEHAVIOR - new UA every request - new socket every request - Connection: close - SOCKS5h via Tor 9050 - external NEWNYM rotation respected
Minimal moving parts. Your system provides Tor. INFILTRATOR provides controlled execution.
SYSTEM REQUIREMENTS - Python 3 - Tor running locally with SOCKS on 127.0.0.1:9050 PACKAGING - Single-file source - PyInstaller --onefile compatible OUTPUT - DaRK.db (external, persistent) - SCRAPES/<domain>/ (mirrored acquisition)