Skip to main content

​
Behavioral Isolation

A Persona represents an isolated behavioral profile, style convention, and rule framework that constrains your active conversations. By establishing distinct archetype profiles, you prevent instruction mixing or conversational context contamination across your localized projects. When a Persona is assigned to a workspace thread, the underlying engine dynamically spins up customized extractor, augment, and read skills tailored to that profile’s specific rule configurations.

​
👤 The Personas Dashboard

The Personas main view lists all active behavioral profiles configured on your local machine.
MemoryBox Personas tab view

The system Personas control dashboard displaying configured active profiles.

​
Default Elements

  • Active Profile Grid: View pre-configured default archetypes or custom profiles you have initialized.
  • Creation Trigger: Clicking the clear ”+ Create Persona” card initializes the isolated creation sequence.

​
🛠️ Initializing a New Persona

When creating or modifying a profile, the system provides an explicit text-field parameters card to dictate reasoning constraints.
Creating a new persona parameter options

The centralized configuration parameters window for initializing a persona.

​
System Configuration Inputs

  1. Name: Declare an explicit string name representing the operational role (e.g., Technical Architect, Data Analyst, Content Strategist).
  2. Summary: Input the raw procedural guidelines, background criteria, and performance parameters the AI model must follow. This field directly informs the system’s dynamic prompt engineering pipeline, which distills your instructions into structured, layered prompt templates that govern the model’s reasoning processes.
  3. Avatar: Select a visual badge to represent the profile across the interface and connected endpoints

​
🎨 Avatar Personalization Matrix

Clicking the avatar profile icon inside the creation panel opens an overlay modal to select or modify visual indicators for your profile card.
Persona avatar image picker overview

Selecting a dedicated visual avatar indicator inside the creation interface.

​
📊 Core Behavioral Traits Matrix

As you interact with the platform, the system maps your global traits against a foundational matrix. These traits are evaluated globally across your local environment, offering an automated discovery layer with specialized confidence scores to track, refine, or sideline your professional defaults.
MemoryBox All Traits side panel drawer view

The global automated discovery library side-drawer panel tracking learned system behavioral trends.

If desired, you can manually pin specific traits to your profile configuration to enforce hard guardrails or explicit performance priorities for the model. The platform actively categorizes these eight parameters to ensure a continuous, deeply personalized agent interface:
Local Store Enforcement: All calculations, discovered confidence scores, and profile configurations remain stored in your local Memory Store Location, maintaining absolute security and sovereign control over your behavioral parameters.

Next Step: Phase 3 — Sandbox Architecture (My Spaces)

With your behavioral personas and traits initialized, proceed to partition your local data assets and profiles into project-specific workflow environments.