Scanning & Stealth
Scanning is how you learn a target, and it is also the first thing that can give you away. This page covers how detection works and how to route around it. For the practical trace loop, see Staying Anonymous.
Nmap and IDS detection
Section titled “Nmap and IDS detection”Two things can see you
Section titled “Two things can see you”- Host IDS: a tier 5+ box running its own intrusion detection. Only that box watches, and only from tier 5 up.
- LAN appliance: a SIEM, IDS, IPS or NDR on a gateway’s LAN. It watches every host on that segment from tier 2 up, even a soft workstation. If the host IDS misses you, the segment appliance still rolls.
Tier 1 to 3 boxes with nothing on the LAN do not watch scans at all.
Scan type is the dial
Section titled “Scan type is the dial”Your scan type multiplies the detection chance:
| Flag | Scan | Multiplier | Trade-off |
|---|---|---|---|
-sT |
TCP connect (default) | x1.0 | Loud |
-sS |
SYN stealth | x0.4 | Quieter, 1.5x slower |
-sN |
NULL scan | x0.1 | Quietest, 2x slower, no versions |
What the odds actually are
Section titled “What the odds actually are”A host IDS starts near 25% and climbs 5 points per tier, then the multiplier applies:
| Host IDS | -sT |
-sS |
-sN |
|---|---|---|---|
| Tier 5 | 50% | 20% | 5% |
| Tier 6 | 55% | 22% | 6% |
| Tier 9 | 70% | 28% | 7% |
A LAN appliance is lower: near 15% plus 3 points per tier before the same multiplier, so a tier 4 segment reads about 27% / 11% / 3% for -sT / -sS / -sN. In short, -sS roughly halves -sT’s risk and -sN cuts it to a tenth. The price is time, and NULL throws away your version detection.
What a trip costs
Section titled “What a trip costs”When a scan is flagged:
- An investigation opens against you. On a proxychains route it opens against your exit node instead, not your real address.
- Heat rises: +0.15 for a host IDS, +0.10 for a LAN appliance.
- The box logs an
IDS ALERTin its syslog, not a routine notice. - Results degrade: about 40% of ports show as filtered, and versions are suppressed.
Your terminal tells you which happened: flagged, a near miss, or a clean pass.
Play pattern
Section titled “Play pattern”- Confirm live and see ports:
nmap -sN <ip> - Pull versions at reduced risk:
nmap -sS -sV <ip> - Route through proxychains first, so any flag lands on your exit node.
- Save
-O,-p-and aggressive scans for boxes that are not watching.
Honeypot servers
Section titled “Honeypot servers”Some networks plant a decoy server whose only job is to catch you. It looks like a normal box in /etc/hosts and on the switch, with a plain server name, so you will not spot it by the name. You spot it with nmap.
Honeypots sit on mid to high tier networks (tier 3 and up), and the security-minded sectors run them most: government and corporate networks are the ones to distrust. There is at most one per network, and its name and type change from network to network, so you cannot learn to guess the next one.
The tell is in the fingerprint
Section titled “The tell is in the fingerprint”Scan it with nmap -sV -O and read the output. A honeypot cannot help giving itself away:
| Tell | What you see |
|---|---|
| Two operating systems at once | Service Info: OSs: Windows, Linux (no real box runs both) |
| nmap says it out loud | Device: honeypot? |
| An impossible service mix | IIS and Samba, or MSSQL and MySQL, on one host |
| A low-confidence OS plus a note | the TCP/IP stack is inconsistent |
No real server runs Windows IIS and Linux Samba and two databases on one host. When the services argue with the operating system, it is a trap.
This is the network honeypot, a whole decoy box. It is not the same as the file honeypots (canary loot files marked INTEL-BEACON-) that corporate and government entities plant after you provoke them; those are covered in the Overview and Free Roam.
Proxychains
Section titled “Proxychains”Every command you run leaves your address in the target’s log. Proxychains puts one or more compromised gateways in front of you, so the log records the exit node instead. A trace still walks the chain from the target backwards, but each hop is a full log-clean-and-disconnect problem before the walker reaches you.
Commands
Section titled “Commands”| Command | What it does |
|---|---|
proxychains --add <ip> |
Push an exploited gateway onto the chain |
proxychains --remove <ip> |
Take a specific hop back out |
proxychains --list |
Show the chain, plus Tor status |
proxychains --clear |
Drop every hop (no session cost) |
proxychains --check |
Test each hop, flag any expired sessions |
proxychains --tor |
Panic-button Tor relay (see below) |
What counts as a hop
Section titled “What counts as a hop”- It has to be a gateway. LAN nodes are not internet-routable and are refused.
- It has to be exploited. You need a live session on the box; cracked SSH creds count.
- No duplicates. The same IP twice is refused.
- The chain cap is 20 hops. Two or three is the sweet spot.
Chain order is add-order: the first hop added is the first out of your machine, the last hop is the exit that touches the target. Remove and re-add to resequence.
The catch: every hop logs too
Section titled “The catch: every hop logs too”Each proxy is a real box, and each hop leaves a session log on it. A chain buys you time, not silence. When you finish, walk the chain in reverse and run logcleaner --ip <your_ip> on each hop. The surgical form takes no dual-track penalty; --all costs a 1.5x trace-speed boost on every box you use it on, so avoid it here.
Check before you fire
Section titled “Check before you fire”Sessions expire. A cold hop that lost its session silently breaks the chain and your traffic drops. --check pings each hop and marks the dead ones:
OK 185.220.101.4 - session activeFAIL 51.158.107.22 - session expiredRe-exploit the dead hop or --remove it before the next command.
Tor relay (--tor)
Section titled “Tor relay (--tor)”proxychains --tor engages a 3-hop Tor circuit (guard, middle, exit) on top of your existing chain. It is a panic button, not a stealth layer:
- Trace algorithms run at 30% speed for about 120 seconds.
- Costs +0.05 heat per activation (exit nodes are surveilled).
- 10-minute cooldown after each circuit; rate-limited to prevent traffic-correlation attacks.
- Re-running
--torwhile a circuit is live shows the active hops and remaining time.
Use it when a trace is closing fast and you need 120 seconds to disconnect and clean up. It stacks with your existing hops, and with VPN and LTE failover.
Recommended flow
Section titled “Recommended flow”- Own two or three gateways in different sectors as your standing chain.
proxychains --addthem before any noisy work (hydra,sqlmap,worm,download).--checkbefore every major action.- When a trace goes red,
--torand disconnect. - Reverse the chain and
logcleaner --ip <your_ip>on each hop. - Disconnect all sessions on the hops. Dangling sessions on your chain are what an investigator subpoenas next.
