# VG-Environment — Project Principles

Hinweis:
Die grundlegende Zieldefinition und das Erkenntnisinteresse von VG-Environment sind in `docs/architecture/project-philosophy.md` beschrieben und bei allen Architektur-, Implementierungs- und Analyseentscheidungen vorrangig zu berücksichtigen.

## Core purpose

VG-Environment is a local evidence and runtime reconstruction environment for consent-related browser activity.

The primary project goal is NOT large-scale telemetry collection or abstract tracking analytics.

The primary goal is:

To technically reconstruct what was actually observed after a concrete consent decision inside the browser runtime.

Core user question:

"What exactly happened technically after I was shown a consent dialog and made a specific consent decision?"

VG-Environment therefore focuses on:
- technically observable runtime behavior
- consent-state reconstruction
- request observation
- GVL/version preservation
- evidence-oriented reproducibility
- human-understandable technical reconstruction

---

## Evidence-first philosophy

VG-Environment prioritizes:
- observable technical reality
- reproducibility
- transparency
- evidence preservation
- explainability
- low interpretation bias

The project must avoid:
- black-box logic
- AI-generated assumptions
- speculative conclusions
- hidden heuristics
- non-reproducible scoring systems
- unverifiable classifications

The browser runtime is treated as the primary source of truth.

---

## Consent-centered architecture

VG-Environment is fundamentally consent-centered.

The project is NOT primarily:
- request-centered
- statistics-centered
- telemetry-centered
- crawler-centered

Requests are treated as:
observed technical follow-up activity related to concrete consent states.

The central reconstruction unit is therefore:
a concrete consent event and its technically observable consequences.

---

## Observe vs Analysis

VG-Observe is responsible for:
- runtime visibility
- technical observation
- evidence capture
- temporal reconstruction
- technical correlation visibility
- human-understandable evidence presentation

VG-Observe may:
- show technical facts
- show observable relations
- show temporal ordering
- show observed request activity
- show observed consent propagation

VG-Observe must NOT:
- speculate
- classify risk
- infer motives
- generate legal conclusions
- claim causality
- invent behavioral models
- silently interpret observations

VG-Analysis is responsible for:
- interpretation
- discrepancy analysis
- pattern analysis
- hypothesis generation
- deeper correlation models
- later analytical evaluation

The separation between Observe and Analysis is a core architectural rule.

---

## Human-understandable evidence

The project must not devolve into:
- raw database dumps
- table graveyards
- developer-only telemetry views

Technical evidence must remain understandable to humans.

Preferred UX principle:
- first orientation and understandable context
- then technical detail
- then raw evidence/debug data

Raw technical detail remains important,
but must not dominate the primary user experience.

---

## Architectural direction

VG-Environment intentionally prefers:
- small modular structures
- explicit logic
- low hidden complexity
- direct inspectability
- Vanilla JavaScript
- minimal runtime dependencies
- local-first workflows

The project intentionally avoids unnecessary:
- framework complexity
- bundler dependency
- hidden build pipelines
- runtime abstraction layers
- architecture inflation

Long-term maintainability and inspectability are more important than trend-driven architecture decisions.

---

## Evidence integrity

Evidence integrity has priority over convenience.

The system must avoid:
- silent evidence modification
- destructive cleanup logic
- hidden data rewriting
- opaque processing pipelines

Stored observations should remain reproducible and explainable whenever possible.

The project is designed as a technical reconstruction environment,
not as an automated decision engine.