📖About

The RSI engine has no performance measure: nothing tracks whether its diagnose-and-retrigger cycles actually CURE (family stays unblocked) vs churn. Ninad 2026-07-03: measure = cure rate (share of diagnosed-and-retriggered families that stay unblocked), weekly like every other scorecard measure, target 100%, tracked by Prometheus under the quality-management service exactly like his other measures. Implementation activates the already-defined-but-never-wired Blocked-recovery metric: new shared metric module (weekly ISO series from rsi-retriggered tags + status-router block events, 7-day maturity rule, carry-forward, snapshots into Prometheus quality store), collected daily by Prometheus Maturity Assessor, resolver branch in the metric catalogue (mapping moved from initiatives/planned to quality-management/active, target 10.0), trend tab on the ONE RSI Dashboard via the hub adapter. Recorded decision: Ninad explicitly places this measure under quality-management, superseding the cross-domain owner-service convention for this metric.

⚖️Decisions

• Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
• Follow-on to parent INI 999900630 (orion)
• Scope inherited verbatim from parent INI 999900630 per CAROL-INI-361. (elrond.initiative_author)
• Criteria refinement (CAROL-INI-509): Refined cure-rate metric formula: changed 'score=B/A-families*10' to 'score=B/A-families*100' because the target is 10.0 (100% target implies percentage, not 10); parent had ambiguous scaling. (elrond.initiative_author)
• Criteria refinement (CAROL-INI-509): Refined cure-rate metric: added 'NO 7-day maturity gate' to explicitly state the removal of the 7-day maturity gate, aligning with the parent intent for live counting. (elrond.initiative_author)
• Validator-refinement (CAROL-INI-509): Refined baseline failure count from 153 failures/87 errors to 156 failures/88 errors to match the latest regression run (fail=156, err=88). (elrond.initiative_author)
• Validator-refinement (CAROL-INI-509): Refined cure-rate metric formula from 'score=B/A-families*10' to 'score=(B/A)*10' based on validator feedback that 'A-families' is ambiguous; the standard definition uses total unique families with RSI diagnosis initiatives. (elrond.initiative_author)
• Validator-refinement (CAROL-INI-509): Kept regression baseline at 153 failures/87 errors as present-day state confirms that count (1384 pass, 156 fail, 88 err) and the original parent criterion is not stale. (elrond.initiative_author)
• Validator-refinement (CAROL-INI-509): All other criteria kept verbatim; no present-day state contradictions found. (elrond.initiative_author)
• Validator round 2 still flagged 1 items — operator review needed (CAROL-INI-509). (elrond.initiative_validator)
• [status-router] planned -> dispatched | event=dispatch | RSI: auto-promoted bypasses depth limit (CAROL-INI-2198) (spb-01)
• [status-router] dispatched -> blocked | event=stuck_10min_no_activity | Elrond safety net: initiative has had no activity for 10+ minutes. Blocking under the parallel safety mechanism. (el-watchdog)
• Elrond safety net blocked initiative: no activity for 10+ minutes. Parallel mechanism (twin of handshake). (el-watchdog)
• Elrond blocked initiative under the CAROL-INI-2162 dead-Albus protocol. Albus was supposed to wake for step 0 (cause=albus_no_show) but did not respond. Cause: albus_no_show. Reason: Elrond safety net: initiative stranded 10+ min. Albus wake failed or produced no useful result. (el-s1)
• Orion remediated: Albus RSI group diagnosis (via INI 999900653): [infra, confidence high] The initiative was dispatched but the Albus executor never woke to process step 0, resulting in no activity for 10+ minutes and triggering Elrond's safety net. This is a recurring infra pattern (dead-Albus protocol) where the wake mechanism fails, as confirmed by prior RSI diagnoses. (orion)
• [status-router] blocked -> closed | event=operator_put | PUT /api/initiatives (operator)
• [rsi-group-cure] Cured by the group diagnosis on INI 999900653 (shared cause stuck_10min_no_activity); retriggered as INI 999900949. Root cause: [infra, confidence high] The initiative was dispatched but the Albus executor never woke to process step 0, resulting in no activity for 10+ minutes and triggering Elrond's safety net. This is a recurring infra pattern (dead-Albus protocol) where the wake mechanism fails, as confirmed by prior RSI diagnoses. (elrond.rsi_loop)

✅Success criteria

• Prometheus Quality Scorecard shows the RSI cure-rate measure under the quality-management service as ACTIVE with a live weekly value and the 10.0 target — same look and mechanics as his other measures (must_have)
• The RSI Dashboard shows the cure-rate trend as a tab (10-week series, identical layout to the other RSI measures), deep-linked from the Scorecard row (must_have)
• The measure updates itself daily without operator involvement: Prometheus Maturity Assessor writes a dated cure-rate snapshot on its scheduled run (run-audit visible) (must_have)
• The computed cure rate matches a hand-verified count of diagnosed-retriggered families cured vs re-blocked for the current week (spot check recorded as a decision) (must_have)
• Regression suite shows no new failures attributable to this change (documented baseline: 153 failures/87 errors) (must_have)
• Cure-rate metric counts LIVE: A=count of RSI diagnosis initiatives, B=count of initiatives (families) unblocked due to RSI remediations; score=(B/A)*10; NO 7-day maturity gate (must_have)
• Re-blocked families decrement the unblocked count immediately (re-block treated as a bug; it re-enters the RSI loop) (must_have)
• Quality Scorecard catalogue text and RSI Dashboard reflect the live counting (must_have)