Artifact: A Postmortem's Root Cause Becomes a Tracked Debt Item

ex-50 · exercises co-23, co-16 · the follow-up item traces to the exact postmortem line that justified it.

Postmortem excerpt (INC-0104, blameless, per Example 43's rewrite):

Root cause: the ops-toggle script had no built-in guard against re-applying a change that was
already in effect, and no confirmation step surfaced the flag's current state before running.

Tracked debt item, filed the same day:

## DEBT-052: ops-toggle scripts lack a pre-run state confirmation guard
 
Logged: 2026-07-18
Owner: @dev-platform-team
Quadrant: RECKLESS + INADVERTENT (nobody chose to skip this -- it was simply never built)
Source: INC-0104 postmortem, "Root cause" section (quoted above verbatim)
 
Follow-up: audit every ops-toggle script for the same missing guard; add a confirmation prompt to
each (this incident's own toggle script already fixed same-day; the audit covers the rest).

Verify: the debt item's Source line quotes the postmortem's Root cause section verbatim, directly tracing the follow-up item back to the specific line that justified it, satisfying co-23's and co-16's combined rule.

Key takeaway: a postmortem that ends at "root cause identified" loses its own value the moment everyone moves on -- the tracked debt item is what carries the finding forward into the backlog where it can actually get prioritized (Example 29's friction-ranking applies to it too).

Why It Matters: without an explicit trace from postmortem to backlog item, a root cause gets identified, discussed once in a review meeting, and then quietly forgotten -- exactly the kind of un-actioned finding that lets the same failure mode recur under a different incident number.