RGDS v1.3.0 — IND delivery alignment (webinar-grounded decision governance)

[mj3b]

Summary

This PR aligns RGDS with how IND programs actually operate under regulatory, timing,
and dependency constraints, based on real-world IND execution patterns discussed
in the Syner-G IND submission webinar.

The goal is not to add autonomy or automation, but to make critical delivery and
regulatory decisions explicit, auditable, and phase-appropriate.

What changed

Decision model enhancements

• Added optional, governance-controlled fields to the decision log schema to support:
  • risk posture declaration
  • author-at-risk drafting
  • reviewer triage and escalation
  • scope change events
  • dependency and data-readiness tracking
  • publishing lock points
• Expanded program_context.domain to explicitly support ind_submission.

Governance and documentation

• Added a role → decision → artifact matrix documenting how RGDS supports:
  PMs, medical writing, regulatory strategy, CMC, ops, quality, and
  the Principal AI Business Analyst role.
• Updated governance and decision-log documentation to reflect IND-phase realities
  (dependencies, late data, escalation, and fallback planning).

Usability

• Updated decision-log template so teams can use IND-aligned fields immediately.
• Updated README and CHANGELOG to clearly describe the v1.3 scope.

What this does NOT do

• No agent autonomy
• No decision automation
• No replacement of human regulatory judgment

RGDS remains explicitly human-governed, evidence-linked, and audit-ready.

Why this matters

These changes reduce late-stage decision surprise by forcing
risk acceptance, scope changes, and drafting-at-risk decisions to be made
explicitly and early—without compromising regulatory trust.

Validation

All existing example decision logs pass schema validation.