Skip to content

IDE RTM Audit Summary

Status: Internal audit summary Owner: BionicLoop engineering Last updated: 2026-04-07 12:32 EDT

Purpose

Summarize the current audit of the Requirements Traceability Matrix for the IDE software packet.

This document does not replace TraceabilityMatrix.md. It is the reviewer-facing summary of what the RTM currently supports for the IDE software packet and what still requires freeze-time action.

Bottom-Line Audit Result

As of this audit:

  • 0 RTM rows are formal-ready
  • 15 RTM rows are rerun-needed
  • 1 RTM row is deferred (partial scope)

Reason:

  • the formal evidence lane under /Users/jcostik/BionicLoop/Docs/Quality/Evidence/Formal/ is not yet populated with promoted STR bundles for the claimed software baseline
  • most RTM rows still reference working evidence, prose descriptions of tests, or implementation-context notes rather than freeze-ready formal evidence

Bucket Definitions

  • formal-ready: already backed by formal STR evidence suitable for freeze-time claim
  • rerun-needed: remains in current packet scope, but still needs formal promotion/rerun before freeze
  • deferred (partial scope): only part of the mapped row is currently claimed; remaining scope is intentionally deferred

Row-Level Classification

RTM Row Audit Bucket Reason Minimal Next Action
RA-001 rerun-needed Runtime cadence row still relies on working simulation evidence. Promote/re-execute formal runtime + simulation evidence for accepted cadence behavior.
RA-002 rerun-needed CGM gating/degraded behavior is documented but not promoted into formal evidence. Promote/re-execute formal CGM policy evidence.
RA-003 rerun-needed Pump-availability and command-block behavior remains in working/support evidence only. Promote/re-execute formal pump policy evidence.
RA-004 rerun-needed Meal workflow has strong support evidence, but still mixes working simulation and support bug evidence. Promote formal meal workflow evidence for accepted current behavior.
RA-005 rerun-needed Delivery reconciliation/logging row names tests but not formal STR evidence. Promote formal automated evidence for reconciliation/logging behavior.
RA-006 rerun-needed State persistence/relaunch/reset behavior is still summarized as partial. Promote formal persistence/state evidence.
RA-007 rerun-needed Live pump-state/device-surface behavior remains pending hardware-backed closure. Execute/promote formal hardware evidence for the claimed pump-state rows.
RA-008 rerun-needed Logging/telemetry behavior is well-described, but not in formal STR form for freeze. Select the in-scope logging rows and promote formal automated evidence.
RA-009 deferred (partial scope) Current packet only claims local SRS-SEC-001..002 interpretation and TV-SEC-001; cloud/auth rows remain deferred. Keep deferred scope explicit and execute/promote formal TV-SEC-001 only.
RA-010 rerun-needed UI/validation behavior has support evidence, but no formal promoted bundle. Promote formal UI automation evidence for accepted IDE-scope UI rows.
RA-011 rerun-needed Alert behavior is broadly covered in development evidence but not promoted into formal freeze artifacts. Promote formal alert automation/manual evidence for claimed alert rows.
RA-012 rerun-needed BG workflow remains in scope except deferred TV-BG-007; formal evidence is not yet promoted. Promote formal BG evidence while preserving explicit defer for TV-BG-007.
RA-013 rerun-needed Clinical settings/governance rows have strong focused evidence but remain support-level, not formal freeze evidence. Promote formal settings/governance evidence for accepted investigational controls.
RA-014 rerun-needed Algorithm campaign has working package linkage, but not formal freeze-lane closure. Execute/promote formal STR-ALG-001 package for the accepted baseline.
RA-015 rerun-needed Runtime interruption/reconnect behavior still calls out remaining real-device evidence gap. Promote formal runtime evidence and execute required live-device confirmation if claimed.
RA-016 rerun-needed Pending/uncertain meal-request and cancel-delivery behavior is covered in tests, but not formal promoted evidence. Promote formal automated evidence for the accepted current meal lifecycle behavior.

Practical Freeze-Time Implication

The current packet is ready for review, but formal evidence promotion is still pending.

For freeze, the remaining evidence work should be limited to:

  1. formal STR-ALG-001 execution/promotion for the claimed algorithm rows
  2. formal STP-AUTO-001 execution/promotion for the claimed automated local software rows
  3. formal TV-SEC-001 execution/promotion
  4. formal STP-HW-001 or STP-ALERT-001 execution only for live/manual behaviors that remain truly claimed in the current packet

Primary Support References