Skip to content

sirraya-labs/Sirraya_TDF

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

9 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

Trust Decay Framework (TDF)

A confidence-based classification system for verifiable credentials with time-based trust decay, zero-knowledge proofs, and risk-aware adaptive security.

License: MIT Status: Editor's Draft


πŸ“– Overview

Trust is not static. A credential that was trustworthy yesterday may not be trustworthy today. The Trust Decay Framework (TDF) provides a standardized approach for assessing and expressing the current trustworthiness of digital credentials over time.

Instead of binary valid/invalid decisions, TDF introduces a four-tier confidence hierarchy where trust naturally diminishes without ongoing evidence of security and identity consistency. Credentials can be refreshed through cryptographic proofs, heartbeat mechanisms, and risk assessmentβ€”all while preserving privacy through pairwise identifiers and zero-knowledge proofs.

Built on established standards like W3C Decentralized Identifiers (DIDs) and Verifiable Credentials (VCs), TDF enables issuers to create credentials with an initial trust level, and verifiers to evaluate the current trust level based on age, cryptographic strength, proofs, and risk signals.


🎯 Core Concepts

Confidence Levels

Level Name Half-Life Typical Use Cases
Level 4 Maximum Trust 30 days Government services, critical systems
Level 3 High Trust 90 days High-value banking, corporate access
Level 2 Standard Trust 90 days Enterprise applications, retail banking
Level 1 Basic Trust 7 days Anonymous access, low-risk services

Half-Life Explained: The period after which confidence decreases by 50% without refreshment. Higher trust levels require more frequent maintenance because they grant access to more sensitive resources.

Trust Operations

  • Demotion β€” Trust decreases due to time decay or detected risks
  • Promotion β€” Trust increases when valid heartbeats are provided (limited to original issuance level)

Key Mechanisms

Mechanism Description
Time-Based Decay Trust diminishes predictably according to a mathematical formula
Zero-Knowledge Proofs Prove security status without revealing sensitive underlying data
Heartbeats Cryptographic proofs that refresh or restore trust (biometric, proof-of-possession, liveness)
Risk Signals Real-time anomalies that trigger immediate trust reduction
Pairwise Identifiers Unique IDs per relationship prevent correlation across services

πŸ“‹ Specification Structure

1. Introduction

Defines the problem statement, design principles, and usage examples. Trust should not be permanentβ€”credentials require ongoing verification.

2. Terminology

Standardized definitions for confidence levels, half-life, heartbeats, risk signals, zero-knowledge proofs, and pairwise identifiers.

3. Trust Levels

Specifies the four-tier hierarchy with half-lives, proof requirements, and demotion/promotion rules.

4. Credential Format

Extends W3C Verifiable Credentials with confidence level information, assessment policy references, and validity timestamps.

5. Trust Decay Calculation

Mathematical formula for computing current trust level:

DecayFactor = 2 ^ (-(EffectiveAge Γ— CryptoMultiplier) / HalfLife)
  • EffectiveAge = Actual age minus time offset from valid heartbeats
  • CryptoMultiplier = 1 for quantum-resistant algorithms, 5 for quantum-vulnerable
  • HalfLife = Based on initial confidence level

6. Proofs and Heartbeats

Defines required proof types for each trust level and heartbeat mechanisms:

  • No High-Risk Activity (24h) β€” Required for Level 2+
  • Device Consistency (7d) β€” Enables Level 3
  • Device Consistency (30d) β€” Enables Level 4
  • Stable Location (72h) β€” Enables Level 4

Heartbeat types include biometric verification, proof of possession, and liveness checks.

7. Risk Signals

Specifies observable events that may indicate compromised security:

  • Device Change β€” Demote to Level 2
  • Geographic Anomaly β€” Temporary demotion (2 hours)
  • Behavioral Anomaly β€” Require heartbeat (15 min)
  • Concurrent Session β€” Severity-dependent action
  • Velocity Anomaly β€” Severity-dependent action

Severity levels range from Low to Critical, with corresponding actions.

8. Cryptographic Requirements

Defines supported algorithms and quantum considerations:

  • Ed25519 β€” REQUIRED (classical)
  • ML-DSA-44 β€” RECOMMENDED (post-quantum)
  • ML-DSA-65/87 β€” OPTIONAL (post-quantum)
  • Hybrid Modes β€” OPTIONAL

Quantum-vulnerable algorithms incur a 5Γ— decay multiplier.

9. Privacy Considerations

Privacy-preserving mechanisms:

  • Pairwise Identifiers β€” Derived per relationship: HMAC(RootKey, RelationshipContext || ServiceIdentifier)
  • Zero-Knowledge Proofs β€” Reveal only that conditions are satisfied, not the underlying data
  • Nullifiers β€” Prevent replay attacks while maintaining unlinkability

10. Implementation Guidelines

Recommendations for state management, multi-tier caching (L1 memory, L2 local, L3 distributed), policy evaluation, and decision types (Granted, Reauthentication Required, Denied).

11. Security Considerations

Attack mitigations and key rotation protocol (pre-rotation, rotation, post-rotation phases).


πŸ”’ Privacy Features

Pairwise Identifiers

Instead of a global identifier, each relationship uses a unique derived identifier. The same holder interacting with different services appears as different entities, preventing correlation without the root key.

Zero-Knowledge Proofs

  • NoHighRisk24h β€” Proves no high-risk alerts exist without revealing which alerts
  • DeviceConsistency β€” Proves device stability without revealing device fingerprint
  • GeoStable β€” Proves location stability without revealing actual coordinates

Nullifiers

Each proof includes a unique nullifier derived from the holder's root key and context, ensuring proofs cannot be replayed or correlated across different contexts.


πŸ” Cryptographic Support

Algorithm Type Quantum Resistance
Ed25519 Classical No
ML-DSA-44 Post-Quantum Yes
ML-DSA-65 Post-Quantum Yes
ML-DSA-87 Post-Quantum Yes
Hybrid Ed25519-ML-DSA Hybrid Yes

Quantum-resistant algorithms receive no decay penalty; quantum-vulnerable algorithms decay 5Γ— faster.


πŸ“Š Trust Decay Example

A Level 4 credential (30-day half-life) issued 45 days ago with no heartbeats:

  • EffectiveAge = 45 days
  • HalfLife = 30 days
  • DecayExponent = -(45 Γ— 1) / 30 = -1.5
  • DecayFactor = 2^-1.5 = 0.3535
  • 0.3535 > 0.25 β†’ Demote two levels
  • Level 4 β†’ Level 2

After a valid biometric heartbeat, trust may be restored or decay reset.


πŸ›οΈ Standards Alignment

The Trust Decay Framework is built on and compatible with:

Standard Relevance
W3C DID Core 1.0 Decentralized Identifiers for subjects and issuers
W3C VC Data Model 2.0 Verifiable Credential structure and proofs
RFC 8032 (Ed25519) Digital signatures

πŸ“„ License

This specification and related materials are licensed under the MIT License β€” freely implementable without restrictions.


πŸ™ Acknowledgments

  • W3C DID Working Group for Decentralized Identifiers
  • W3C VC Working Group for Verifiable Credentials
  • UDNA Community Group for decentralized addressing standards
  • Sirraya Labs for open-source commitment

Trust Decay Framework β€” Making digital trust dynamic, privacy-preserving, and quantum-ready.

Version 1.0.0 β€’ Editor's Draft β€’ April 2026

About

Trust Decay Framework: A Confidence-Based Classification System for Verifiable Credentials

Resources

License

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages