Autonomy Realms

Reflections


Reflections are AI-generated interpretive layers on your signals, clusters, and shapes. Where analysis extracts structured metadata (title, tags, themes), reflections offer deeper readings — narrative storytelling, self-reflection, archetypal interpretation, and ancestral perspective.

How Reflections Work

When you generate reflections, the AI reads the signal's content and metadata, then writes a reflection through a specific interpretive lens. Each reflection type produces a different kind of reading of the same source material.

Reflections are generated per entity — you can create reflections for individual signals, for clusters (drawing on the signals within), and for shapes (drawing on the clusters within).

Built-In Reflection Types

Four reflection types are built into every realm:

NARRATIVE — a chronological storytelling lens. The AI tells the story of what happened, rendering the signal's content as narrative prose. This is about sequence, causality, and the human experience of events unfolding in time.

MIRROR — a direct self-reflection lens. The AI holds up a mirror, showing you what is present in the content without adding interpretation. This is about clarity, presence, and seeing what is actually there.

SYMBOLIC — an archetypal and mythic lens. The AI reads the content through symbolic, archetypal, and metaphorical frameworks — finding universal patterns, mythic resonance, and deeper layers of meaning beneath the surface.

LINEAGE — an ancestral and community lens. The AI places the content in the context of lineage, intergenerational patterns, community, and the relationships that extend beyond the individual. This is about connection across time and kinship.

Custom Reflection Types

You can define custom reflection types for your realm in Settings > Custom Reflections. Each custom type has:

  • A name — the label for this reflection type
  • A user prompt — the instructions that tell the AI how to interpret content through this lens

Custom reflections appear alongside the built-in types when generating reflections for any entity.

System and User Prompts

Each reflection type uses two prompts:

  • System prompt — defines the AI's role and epistemic framework. System prompts are hardcoded for the built-in types and cannot be edited. They ensure the AI operates within Autonomy Realms' interpretive principles.
  • User prompt — the actual instructions for generating the reflection. For built-in types, these are configurable. For custom types, you write the user prompt yourself.

Visibility

Each reflection has its own visibility setting, independent of the signal it belongs to:

  • Public — anyone can read the reflection
  • Sanctum — only subscribers can read it
  • Private — only you can see it

This lets you keep certain interpretive layers private while sharing others. A signal might be public, but its MIRROR reflection could be private — visible only to you.