IDNA: Identity Compression

Your identity isn't a preference list. It's a pattern.

VALIDATED - Protocol tested across 12 models · 8 providers

The Problem with Preferences

Most personalization systems ask: What do you like?

  • Tone: casual or formal?
  • Length: short or detailed?
  • Style: direct or diplomatic?

This creates a surface profile. A list of settings. A configuration file.

But identity isn't configuration. You don't make decisions by consulting a preferences database. You have patterns - deep structures that shape how you think, speak, and act across every context.

Preferences describe the surface. IDNA encodes the structure.

What IDNA Captures

Tone Signature

Not "professional" or "casual" - your specific frequency: sentence rhythm, ratio of direct to diplomatic, where you soften, where you don't.

Decision Architecture

How you actually make choices: what you optimize for when you can't have everything, your hierarchy of values, where you take risks.

Boundary Patterns

Your invisible lines: what triggers your "this matters" response, where you push back, what you protect even when it costs you.

Trust Logic

How you build and break trust: what earns your confidence, how fast you extend it, what breaks it irreparably.

The Compression

From calibration responses, we extract structure and compress it into a symbolic signature. This isn't a transcript. It's a minimal invariant set - the smallest encoding that preserves identity across transformation.

IDNA:*/[ROLE]:*[WHY]~[HOW]->[WHAT]

Example:
IDNA:*/Architect:*precision~structure->clarity

Why Compression Matters

If you need 1,000 tokens of instructions to "be yourself," you've lost 1,000 tokens of context window. Every conversation starts with overhead.

IDNA is compressive. Minimal structure, maximum signal. The AI doesn't need your life story. It needs your invariants.

IDNA doesn't make AI smarter. It makes AI yours.