==================================================
Sagitta Research Notes
==================================================
Status   : Working notes / non-final
Audience : Internal research and design

Purpose:
--------
This document captures exploratory thinking related to capital continuity,
stress modeling, and allocator policy tuning. These notes are not specifications
and may evolve as systems and doctrine mature.

==================================================
Research Area 1: Continuity Model Assumptions
==================================================

Continuity is treated as a design property, not a runtime guarantee.

Assumptions:
------------
- Capital systems will experience partial failure rather than total collapse
- Not all subsystems fail simultaneously
- Human intervention may be delayed, unavailable, or conflicting
- Data inputs may degrade before systems fully fail

Design priorities:
------------------
- Bounded behavior under uncertainty
- Preservation of core invariants
- Graceful degradation over optimization
- Recovery paths that do not require perfect information

Open questions:
---------------
- What minimum signal set is required to maintain safe operation?
- How long can systems operate under stale or degraded data?
- Which invariants must never be violated, even under extreme uncertainty?

==================================================
Research Area 2: Capital Stress Scenarios (Placeholders)
==================================================

These scenarios are conceptual placeholders used to test assumptions,
not predictions.

Scenario classes under consideration:
-------------------------------------
- Liquidity contraction across correlated assets
- Pricing oracle divergence or staleness
- Volatility regime shifts exceeding historical norms
- Counterparty or venue unavailability
- Governance paralysis or delayed coordination

Stress modeling goals:
---------------------
- Identify invariant breakpoints
- Understand compounding failure effects
- Observe allocator behavior under constraint tightening
- Measure recovery latency rather than peak loss

Open questions:
---------------
- Which stress signals matter earliest?
- How should constraint strictness change under stress?
- When is inaction safer than forced reallocation?

==================================================
Research Area 3: Policy Tuning and Constraint Design
==================================================

Policy tuning focuses on constraint behavior, not performance maximization.

Primary tuning dimensions:
--------------------------
- Position concentration limits
- Liquidity sensitivity
- Rebalance frequency
- Volatility tolerance
- Regime transition thresholds

Design principles:
------------------
- Constraints should be explicit and reviewable
- Constraint interactions should be predictable
- Policy behavior should degrade conservatively
- Policies should avoid hidden optimization pressure

Observed tensions:
------------------
- Responsiveness versus stability
- Diversification versus liquidity depth
- Drawdown avoidance versus recovery participation

Open questions:
---------------
- How conservative should default policies be?
- Should policies self-tighten under uncertainty?
- How much policy discretion is acceptable before review?

==================================================
Research Area 4: Decision Interpretability
==================================================

Decision interpretability is treated as a first-order research concern.

Assumptions:
------------
- Decisions will be reviewed after the fact
- Reviewers may not be the original designers
- Explanation quality affects governance trust

Research focus:
---------------
- Minimal explanation sets
- Separating signal from narrative
- Avoiding post-hoc rationalization
- Aligning explanations with doctrine language

Open questions:
---------------
- What is the smallest explanation that preserves meaning?
- How should uncertainty be communicated?
- When should explanations defer to constraints rather than signals?

==================================================
Research Area 5: Long-Horizon Behavior
==================================================

Sagitta systems are expected to operate over long horizons.

Considerations:
---------------
- Policy drift over time
- Regime bias accumulation
- Silent constraint erosion
- Operator complacency

Research directions:
--------------------
- Periodic policy review