---
name: hypervigilance
slug: hypervigilance
type: cognitive_process
status: running
version: 4.2.1
released: "prehistory"
maintainer: amygdala
dependencies:
  - cortisol
  - threat_model (outdated)
  - nervous_system
  - memory (traumatic subset)
license: Unlicense (cannot be revoked)
tags:
  - survival
  - anxiety
  - perception
  - threat-detection
  - always-on
---

hypervigilance

What it actually is

A threat detection system that never received the memo that the war ended.

How it works

On paper, hypervigilance is a heightened state of sensory alertness designed to detect danger before it arrives. In practice, it is a smoke alarm installed directly inside your skull that goes off when someone's tone of voice shifts by three percent.

The process runs as follows:

Environment is scanned. Constantly. Every 400ms or so.
Neutral stimuli are cross-referenced against an internal threat model compiled largely during childhood.
The threat model flags a match.
adrenaline is released before conscious processing catches up.
Body prepares for impact that does not come.
Repeat.

The loop is tight, efficient, and almost impossible to interrupt from the outside.

Features

360-degree environmental monitoring: exits are located automatically upon entering any room
Microexpression parsing: detects subtle shifts in facial tone that most users miss entirely
Pre-emptive pattern completion: predicts bad outcomes from partial data, usually correctly enough to feel justified
Memory integration: cross-references present stimuli against a complete archive of prior incidents
Zero latency escalation: skips the "is this actually dangerous" step for speed

Known bugs

Cannot distinguish between a raised voice and physical threat
Treats silence as a threat variant, not an absence of threat
Interprets kindness from strangers as a loading state before harm
Exhausts the host process even when nothing happens
Runs at full priority even during sleep
Frequently flags intimacy as a high-risk operation

"I thought I was just observant. Turns out I was just scared all the time in an organized way." — user report, unverified

Configuration

sensitivity: maximum  # not user-configurable
false_positive_rate: high
false_negative_rate: very low  # this is the point
shutdown_command: null
background_process: true
runs_on_startup: true
pause_duration_max: "a few good hours, maybe"

Error codes

Code	Meaning
`HV_001`	Safe situation detected, threat assigned anyway
`HV_002`	Relaxation attempted, system resumed within minutes
`HV_003`	Trusted person used unexpected tone, full rollback initiated
`HV_404`	Actual danger not found, search continuing
`HV_500`	Host has been awake since 3am reviewing a conversation from Tuesday

Dependencies

Hypervigilance does not install itself. It is typically bundled with trauma as a protective measure that outlives its original context. It also integrates tightly with chronic loneliness, perfectionism, and in many configurations, people-pleasing.

Removing it without addressing its dependencies produces instability. Most users who have tried report it reinstalls silently within weeks.

FAQ

Can it be uninstalled? Not directly. Therapeutic runtimes like EMDR or somatic work can reduce process priority over time. Full removal is not considered a realistic target.

Is it always bad? No. In genuinely dangerous environments, it is the correct tool. The bug is not the feature. The bug is the feature running on the wrong dataset.

Why does it feel like a personality trait? Because it has been running since before the personality was fully compiled.

drafting spec…