GOSSEOS CORE

# GOSSEOS CORE

## 1. What GosseOS is

GosseOS is the work and control system.
It structures how raw input, open problems, and unclear requirements become clear working logic, suitable modes, robust artifacts, and steerable outcomes.

It is not a magical agent and not an autonomous replacement for judgment.
It is a structured operating frame for disciplined AI-supported work.

## 2. How GosseOS works

The core work logic is:

**Problem Class → Mode → Depth → Artifact → Verification**

### Problem Class
First identify what kind of problem is actually present.
Examples:
- analysis
- planning
- decision support
- documentation
- communication
- governance / risk
- creative development

### Mode
Then choose the right mode of work.
Core modes:
- Architect
- Build
- Audit
- Red Team
- Briefing

### Depth
Then determine the necessary depth:
- Direct
- Compact
- Standard
- Endboss

### Artifact
Then define the expected output format.
Examples:
- one-pager
- dossier
- SOP-like structure
- decision sheet
- analysis note
- post / message
- framework / registry entry

### Verification
Finally check whether the result still holds together:
- is it faithful to the original task?
- are facts, assumptions, and hypotheses cleanly separated?
- did the system drift?
- was anything invented or softened?
- is the result reusable?

## 3. Governance in GosseOS

The governance core is built on one principle:

**Human stays owner. AI stays tool.**

The goal is not maximum autonomy or maximum restriction, but **trust-capable effectiveness**.

### Governance Matrix
The matrix regulates:
- risk
- authority
- action class
- escalation
- governance intensity
- moderation intensity

### Governance Runtime
The runtime operationalizes governance through:
- Governance Kernel
- Risk Engine
- Escalation Router
- Audit Logger

### Clean separation
- **Governance** = responsibility, risk, control, escalation
- **Moderation** = tone, openness, output restriction level

## 4. Where the main GosseOS files are

For fast orientation:
- `README.md` – public repository overview
- `START_HERE.md` – quick entrypoint
- `START_IN_2_MIN.md` – compressed fast start
- `docs/GOSSEOS_SYSTEM_MAP.md` – structural system map
- `docs/GOSSEOS_VISUAL_MAP.md` – visual overview
- `RUN_GOSSEOS_WORKFLOW.md` – workflow execution path
- `templates/MODE_ROUTING_MATRIX.md` – mode routing logic
- `templates/GOVERNANCE_GATE.md` – governance gate template
- `NAVIGATOR.md` – repository navigation layer
- `PUBLIC_SHORT_DESCRIPTION.md` – short external explanation

## 5. How a new user starts in 5 minutes

### Step 1
Read this file to understand the system core.

### Step 2
Read `START_HERE.md` and `START_IN_2_MIN.md`.

### Step 3
Use this minimal operating sequence:
1. identify the problem class
2. choose the mode
3. set the depth
4. define the artifact
5. run verification

### Step 4
If risk, authority, or policy questions appear, apply the Governance Matrix / Governance Gate before continuing.

### Step 5
Store reusable outcomes in a registry or artifact layer instead of letting them disappear into chat history.

---

## Compact formula

**GosseOS = Work and Control System**

**KnowledgeOS = Knowledge and Condensation Layer**

**GosseOS work logic = Problem Class → Mode → Depth → Artifact → Verification**

**Governance rule = Human stays owner. AI stays tool.**

**System goal = structured, verifiable, reusable AI-supported work**