The ADHD-HTB playbook: hacking the brain that hacks the box

The bottleneck on HackTheBox was never the boxes. It was the eleven minutes between deciding to hack and actually typing nmap. ADHD doesn’t tax the work. It taxes the transition into the work. So the real exploit target was never the machine. It was the activation energy, and this whole playbook treats my own attention as the box to pwn.

What this is#

Ten methods, each built to bypass one specific failure mode of the ADHD brain instead of trying to “fix discipline.” They fall into five categories, because each one is a different lever:

• Doomscroll-flipping turns the dopamine slot machine against itself: a Finsta-of-Doom feed of enumeration prompts, swipe-to-pwn reps. If the thumb is going to scroll regardless, point it at attack paths.
• Kinetic hacking chains cognition to the body. Squat while the Nmap scan runs. Pace the attack tree out loud. The fidget is the focus, not a distraction from it.
• Gamification bolts a second, faster scoreboard over the slow one: Ironman roguelike runs, a Forfeit Heist where stalling costs you something that actually stings.
• Friction-bypassing micro-commits shrink the doorway to something the brain can’t talk itself out of: a 90-Second Terminal Tax, a Loaded Gun left chambered from last session so there’s no cold start to dread.
• Sensory anchoring gives the state a physical handle: a 140 BPM pwn liturgy, a texture totem that means we are working now.

It closes on the meta-protocol: how to stack methods without burying yourself, and how to ride out the RSD spike when a box stonewalls you and the lizard brain files it as personal failure instead of what it actually is, a closed port.

The whole thing is the attacker’s mindset swung around to face inward. Find the trust boundary the system is hoping you’ll skim past, except this time the system is me and the boundary is the exact moment I open a new tab instead of a shell.

What got built#

Two of the ten concepts graduated from idea to working tool in a single sitting:

• Swipe-to-Pwn Anki deck (.apkg, 54 cards). Decision-format flashcards: the front is a signal you’d actually catch on a box, the back is the attack path it implies. Port and service enumeration, web-app tells, Linux and Windows/AD privesc, plus a handful of mindset cards. It’s pattern recognition compiled down into spaced repetition, so the lookup hardens into a reflex.
• htb-operator.bashrc, a paste-ready shell config. A pwn launcher that is the Terminal Tax, recon functions with squat reminders wired in, a note/next system for leaving the Loaded Gun chambered, and resume for a zero cold start. Plus quick helpers and tmux-resurrect notes, so a dropped session doesn’t reset the activation cost back to maximum.

Open thread#

Still on the bench: a one-shot installer that checks for and grabs the missing tooling (ffuf, gobuster, seclists, evil-winrm, and friends) so a fresh box never stalls on a command not found. Which is just the same thesis wearing a different hat. Every missing tool is a transition cost, and transition cost is the only enemy in the room.

The honest question this whole project is really testing: does engineering the environment beat trying to engineer the willpower? My bet is the environment wins every time, which is why this lives next to the quest log and not in some productivity-shame journal. Reps logged, not resolutions made.