Skip to content

README.md

Latest commit

 

History

History
62 lines (42 loc) · 2.41 KB

README.md

File metadata and controls

62 lines (42 loc) · 2.41 KB

Examples

These are illustrative RGDS decision records (v2.0.0 whitepaper-aligned) intended to show how the Decision Log schema is used in practice across common phase-gate scenarios in regulated delivery.

Each example represents a single, concrete decision with explicit ownership, evidence linkage, risk posture, and governance controls.

v2.0.0 alignment note

Examples include:

  • mandatory options_considered
  • evidence completeness per evidence item (evidence.evidence_items[].completeness_state)
  • explicit risk_posture and structured residual risk (risk_assessment.residual_risk_items)
  • structured AI assistance disclosure when ai_assistance.used=true

Canonical examples (start here)

The following examples are considered canonical because they collectively demonstrate the full RGDS operating model across common decision outcomes.

  • rgds-dec-0001.json — Canonical conditional_go
    Go decision with explicit, owned conditions, governance approvals, and clearly bounded follow-up actions.

  • rgds-dec-0002-no-go.json — Canonical no_go
    Defensible refusal with documented rationale, risks, and a defined re-entry path.

  • rgds-dec-0003-defer-required-evidence.json — Canonical defer / abstain
    Decision deferred pending required evidence, with explicit gaps and re-review criteria.

  • rgds-dec-0004-regulatory-interaction.json — Canonical regulatory interaction / escalation
    Pre-IND or agency-facing decision framing, including questions, strategy, and governance rationale.

  • rgds-dec-0005-ind-conditional-go-author-at-risk.json — Canonical IND-style conditional_go
    IND readiness decision demonstrating author-at-risk drafting, reviewer triage, and publishing lock points.

  • rgds-dec-0006-ai-assisted-conditional-go.json — Canonical AI-assisted conditional_go
    Conditional-go decision demonstrating bounded AI assistance with explicit disclosure, preserved human authority, and full auditability.

Notes on AI usage in examples

rgds-dec-0006-ai-assisted-conditional-go.json is the only example in this repository that demonstrates AI-assisted decision preparation.

All other examples are fully human-authored and remain valid demonstrations of the RGDS operating model without any AI involvement.

This is intentional: RGDS is designed to be AI-optional, and all decisions must remain defensible in the absence of AI assistance.