[Research] Repair-triggers vs Invalidation: Temporal semantics of defeasibility

## Research Question

CLAIR v1 has **invalidations**: conditions that trigger reconsideration of a committed Belief.

INTENT proposes **repair-triggers**: conditions that trigger re-negotiation during the Intent phase.

Are these the same concept at different times, or fundamentally different mechanisms?

## Comparison

| Feature | Invalidation (CLAIR) | Repair-trigger (INTENT) ||---------|---------------------|------------------------|
| Timing | Post-commitment | During negotiation || Action | Mark belief suspect | Re-enter negotiation || Actor | Monitor/auditor | Any participant || Persistence | Stored with Belief | Ephemeral to conversation |
## Key Distinction

**Invalidation** says: "This belief we agreed on—it's now questionable."

**Repair-trigger** says: "We haven't agreed yet, and here's why we might not."

## Research Questions

1. Can repair-triggers be "promoted" to invalidations upon commitment?
2. Should invalidations be "demoted" to repair-triggers if the Belief is reopened?
3. Is there a unified defeasibility mechanism that works across both phases?

## Formalization Goal

Define a temporal logic where:
- Pre-agreement: Conditions trigger repair (re-negotiation)
- Post-agreement: Same conditions trigger invalidation (retraction)
- Transition rules govern movement between phases

---
*Research thread: Defeasibility across negotiation and commitment*