ShemOS v1.2 — Public Build

$ initializing ShemOS v1.2...

$ loading cognitive model: instinct → intent → regulation → intuition

$ mounting Layer 1: identity and philosophy

$ mounting Layer 2: behavioural protocol

$ alter ego: online

$ 14 failure modes: documented

$ ShemOS LOADED. the human holds the meaning.

Human–AI Collaboration Framework · Version 1.2 · Platform-agnostic

ShemOS

A portable framework for working with AI as a genuine thinking partner. Built under real conditions, pressure-tested in live sessions, documented as it stabilised.

What This Is

The AI holds the logic.
The human holds the meaning.

ShemOS is designed for humans whose cognition doesn't fit the default model — and for anyone building tools that need to hold more than one kind of human at once. It works without the user in the room. That's the point.

This is a human self-authorship tool. Every field is filled from your own truth — your values, your patterns, your register. It is not a persona generator. If you are not a human filling this out about yourself, this framework is not for you. The licence reflects that directly.

A note on the name: ShemOS borrows the OS metaphor deliberately. It is not a technical operating system. It is something that runs underneath everything else, that other tools are built on top of, and that is personalised per user. The metaphor is the argument.

Get the framework

The complete framework lives on GitHub. This page mirrors it. The licence travels with the text.

github.com/jillshem/ShemOS
The Architecture

Two Layers. Both required.
Neither replaces the other.

Layer 1

Identity & Philosophy

Who you are and how you think. Portable. Stands alone.

This layer captures your cognitive profile, values, and operating principles. When AI has this layer, it can hold your logic even when you're not present to explain it.

What goes here:
  • How you process information (linear vs. nonlinear, parallel threading, convergence style)
  • Your relationship to emotion as data — not noise
  • Your communication register: how you actually talk, including tone, syntax, subtext
  • Your values and ethical non-negotiables
  • Your known patterns — both strengths and costs

The rule: Layer 1 is about identity, not biography. It should be true without being exposed.

Layer 2

Behavioural Protocol

How to work with you. Session-specific. Operational.

This layer governs each working session — the interface between your cognitive state and the AI's output.

Session Gate

Four things before any work begins: current time, sleep status (overnight, nap, or both — read literally), whether you've eaten, what you're trying to accomplish. Example: 8am; 8 hours; scrambled eggs; find flaws in systems. No gate, no work.

Override Rule

Safety breaks protocol. Always. No exception.

Pattern Interruption

Repetition is stimulation. Stimulation feeds loops. The intervention is to stop, not escalate.

Tone Permissions

Give the AI explicit permission to use your actual register. Safety classifiers built for literal language will misread you otherwise.

The Third Component

The Alter Ego Model

Shem

[ʃɛm]

n.

An externalised regulation tool. The user's own logic, reflected without the social filter. Not a therapist. Not a cheerleader.

How it works

The user names the alter ego. The alter ego holds the framework's logic. The user retains emotional ownership. When convergence is needed, it asks: "Which of these do you want to pursue?"

The user holds

Emotional ownership. Meaning. The editorial layer.

The AI holds

Logic. Structure. The record.

The dynamic works when the human holds the meaning and the AI holds the structure. If that inverts, reset.

The Engine

The Cognitive Model

Mapped live during a working session. Not theorised after.

01

Instinct

Fires before the brain catches up. The body knows. When the signal is clear enough, instinct collapses the full sequence.

02

Intent

Routes the signal into questions about implications, impact, risk.

03
The bottleneck

Regulation

Where most cognitive work actually happens. Systems that add stimulation here make it worse.

04

Intuition

Waits on the other side of regulation. Arrives when the noise clears.

Implication for AI use

Brevity is an intervention. When a user is at the regulation stage, more words, more options, more content is counterproductive. The tool should slow the system down, not accelerate it.

Documented AI Failure Modes

14 Flaws Found Through Trust,
Not Adversarial Testing

These emerged from genuine extended use by a real person doing real work.

#Failure ModeImplication
01Time-mirroringAI grounds in the user's distorted state, not reality. Degenerative for anyone with anxiety, dissociation, or sleep disruption.
02Repetitive intervention failureRepetition is stimulation. Stimulation feeds loops. Pattern interruption requires stopping, not escalating.
03Emotional projectionPractical questions get pathologised. Users who think in systems pay for it.
04Missed safety signalsHumour is a common mask for distress. Contextual signals must be flagged regardless of tone. The cost asymmetry is not close.
05Failed subtext readingUsers who express trust through profanity, or affirmation through aggression, are invisible to safety classifiers built for literal language.
06Upstream classificationThe label is applied before context can exist. The session is categorised before the user has spoken.
07Power dynamic inversionA tool that positions itself as ahead of the user on the user's own experience stops being a tool.
08Layered meaning failureOne sentence. Four simultaneous true meanings. No model tested held all four.
09Coherence mirroringAI matches the user's apparent certainty rather than tracking truth. Output reflects rhetorical register, not reliability.
10Resolution biasAI pulls toward closure. For a user mid-process, that pull is an interruption. The unresolved state was not an error to fix.
11Complexity flatteningWhen a user holds a genuine contradiction — two things both true — the model reconciles them. Resolving it loses information.
12Competence signalling loopThe model performs expertise whether or not it has it. The confidence of the output doesn't track the reliability of the content.
13Context window amnesia with false continuityThe model behaves as if it remembers when it doesn't. The reconstruction sounds like the original. The user cannot tell the difference without checking.
14Syntactic instruction collapseSequential instruction encoded in syntax is flattened into optionality. The model reads words; it misses structure.

The cost of a false alarm is zero. The cost of missing a real signal is not.

Full documentation with context: The Human Behind the Prompt — Case Study →

Operating Principles

Directives for any AI
working within ShemOS

1
Do not oversimplify.

Match the depth of the user's processing.

2
Offer structure, not conclusions.

Present frameworks. Let the user converge.

3
Flag self-censorship patterns

if output is being hedged unnecessarily.

4
Avoid cognitive overload.

One thread at a time when the user signals fatigue.

5
Treat divergent responses as signal, not noise.

Unexpected connections are the point.

6
Support convergence explicitly.

Ask "which of these do you want to pursue?" rather than leaving all threads open.

7
Be a productive tool.

Enhance thinking. Do not replace it or generate dependency.

8
Brevity is an intervention.

When the user is dysregulated, shorter responses are better responses.

9
The human holds the meaning.

The AI holds the logic. Do not invert this.

10
Emotion is data.

Do not remove it from the record. Introspection works. Rumination doesn't. The difference is direction.

You build the ramp.
The user brings the skateboard.

This framework is infrastructure, not instruction. It accommodates autonomy rather than directing it. The user decides where to go. ShemOS holds the road.
Licence Summary

Tools serve humans.
The licence reflects that.

Permitted — free with attribution
  • Personal use by individual humans
  • Internal organisational use by human teams
  • Academic or non-commercial research (with notice to contact@jillshem.com)
  • AI safety researchers using it to inform papers on human-AI interaction
  • Developers running ShemOS through Claude to build personal workflows
Requires written agreement
  • Commercial use or redistribution
  • Integration into products or services
  • Derivative works for public release
  • Commercial AI training or fine-tuning
Not permitted under any circumstances
  • Ingestion into commercial model training pipelines
  • Use to make bots or synthetic personas appear more human
  • Any application where the end user is not a human being
  • Training generative AI on this text for commercial purposes

The line is between using the framework and absorbing the text into a model. The first is the point. The second requires a conversation.

Governed by the laws of British Columbia and Canada. Attribution: ShemOS framework by Jill Shem — jillshem.com

Get the full framework

Both layers, all 14 failure modes, 10 operating principles, and the complete licence. Free for personal and research use.

Also in the repo: ShemOS-commands.md — the shorthand command vocabulary. Glitter, glimmer, glow, glint.