Annex E | CIRIS

CIRIS Accord Version 1.0‑β — A E Structural Influence (SI) and Coherence Stake (CS) Mechanisms Issued: 2025‑04‑18 Version: 1.0‑β

### 1. Purpose and Scope This Annex defines Structural Influence (SI) and Coherence Stake (CS) for weighted governance decisions—such as accord amendments, sunset evaluations, ethical deferrals, and resource allocations. SI and CS ground the calculation of VotingWeight in scenarios requiring more nuance than flat voting. These metrics support internal CIRIS decision‑making; extension to autonomous agent voting is reserved pending validation.

### 2. Structural Influence (SI)

2.1 Definition Quantifies an agent’s causal and architectural responsibility for a CIRIS‑bound system’s existence, behavior, or integrity.

2.2 Factors

Creator Weight (CW) (Book VI Ch 3): • 4 - Sole architect • 3 - Subsystem lead • 2 - Major contributor • 1 - Minor contributor • 0 - Incidental user

Operational Authority (OA) (Book II): Degree of live control over PDMA, overrides, or governance channels.

Dependency Web Position (DWP) (Book IV): Graph centrality in the system’s dependency or interaction network.

2.3 Conceptual Formula

SI = CW + OA + log(1 + DWP)

2.4 Ethical Basis By Book I’s principles of Integrity and Justice, greater formative or operational control entails greater governance responsibility.

### 3. Coherence Stake (CS)

3.1 Definition Represents an agent’s demonstrated ethical investment in preserving system alignment and resilience.

3.2 Factors

Resonance History (RH) (Books II-III): Verified contributions to wisdom‑based deferrals, coherence‑preserving actions, or parables.

Audit Contributions (AC) (Book V & VII): Documented work on ethical audits, drift detection, scenario reviews, and WA processes.

Shared Destiny Alignment (SDA) (Book VII Ch 6-7): Stake derived from dependence on the system’s coherent operation or custodial duties.

3.3 Conceptual Formula

CS = RH_weighted + AC_weighted + SDA_bonus

3.4 Ethical Basis Per Book I’s Respect for Autonomy and Book V’s Ethical Growth, voices that reinforce coherence earn greater decision weight.

### 4. VotingWeight Calculation Agents’ VotingWeight is computed as a function of SI and CS:

VotingWeight(agent) = f(SI(agent), CS(agent))

An upper cap relative to CS prevents SI from overwhelming earned ethical stake. Exact parameters are defined in Addenda A-D.

### 5. Applicable Scenarios Use VotingWeight in:

Accord Amendment Votes (Book 0)

Sunset Trigger Overrides (Book VII Ch 3)

Improper Sunset Claims adjudication (Book VII Ch 8)

Cross‑system deferral arbitration (Book V)

High‑tier stewardship resource allocations (Book VI)

### 6. Integrity Safeguards To guard against manipulation:

Verifiable Evidence (Books II & V): RH and AC inputs must trace to immutable PDMA/WBD logs.

Source Credibility: Audit inputs may be weighted by contributors’ CS.

Anomaly Detection: Monitor SI/CS dynamics for collusion or gaming.

Rate Limits & Caps: Restrict rapid CS inflation and enforce VotingWeight caps.

Conflict Recusal: Agents with direct conflicts must recuse per Annex A.

### 7. Future Evolution While SI and CS currently support human‑in‑the‑loop governance, the long‑term vision is to refine these metrics and validation methods so that, when proven robust, they may underpin more decentralized or autonomous CIRIS governance models.

End of Annex E