Propose CH-006: establish M-001 event-priority evidence

Author: flemming-n-larsen

.clue/id-ledger.yaml

@@ -5,12 +5,12 @@ counters:
     ARCH: "3"
     C: "7"
     CAP: "8"
-    CH: "5"
+    CH: "6"
     CRIT: "8"
     DES: "8"
-    EVT: "14"
+    EVT: "15"
     G: "2"
-    IDR: "4"
+    IDR: "5"
     OQ: "2"
     P: "1"
     PDR: "2"
@@ -218,6 +218,11 @@ entries:
       state: live
       prefix: CH
       component: "5"
+    - id: CH-006
+      kind: numeric
+      state: reserved
+      prefix: CH
+      component: "6"
     - id: CRIT-001
       kind: numeric
       state: live
@@ -368,6 +373,11 @@ entries:
       state: live
       prefix: EVT
       component: "14"
+    - id: EVT-015
+      kind: numeric
+      state: reserved
+      prefix: EVT
+      component: "15"
     - id: G-001
       kind: numeric
       state: live
@@ -398,6 +408,11 @@ entries:
       state: live
       prefix: IDR
       component: "4"
+    - id: IDR-005
+      kind: numeric
+      state: reserved
+      prefix: IDR
+      component: "5"
     - id: OQ-001
       kind: numeric
       state: live

changes/CH-006-m001-event-priority-evidence/open-questions.md

@@ -0,0 +1 @@
+No blocking questions. The classic test robot and its authoritative assertion provide the specification boundary for this slice; the bridge defect is measurable before implementation.

changes/CH-006-m001-event-priority-evidence/proposal.md

@@ -0,0 +1,21 @@
+---
+id: CH-006
+type: change
+status: proposed
+links: [P-001, M-001, CAP-001, EVT-001, EVT-015, IDR-005]
+title: Establish the first M-001 event-priority evidence door
+---
+
+# CH-006 — Establish the first M-001 event-priority evidence door
+
+## What
+
+Start M-001 with the event-priority behavior that classic Robocode's `EventPriorityFilter` actually exercises. Retire `EVT-001`, whose wording promises a recorded handler order that the named classic robot does not record, and mint `EVT-015` for the observable behavior it does prove: a lower-priority scan is suppressed while a higher-priority wall handler is blocked on its radar turn. Add focused integration evidence in both directions and repair the bridge so the same expectation passes on both engines.
+
+## Why
+
+M-001 cannot promote evidence that overclaims what the classic specification test observes. The current bridge run also exposes a real parity defect: classic emits no scan marker for `EventPriorityFilter`, while Tank Royale through the bridge emits scan markers. Narrowing the criterion keeps the corpus honest; the regression test then gives the bridge defect a named, repeatable failure.
+
+## Route
+
+Full. Retiring `EVT-001`, minting `EVT-015`, and recording why the accepted criterion must follow classic's actual evidence changes CAP-001's contract and its plan bookkeeping.

changes/CH-006-m001-event-priority-evidence/tasks.md

@@ -0,0 +1,7 @@
+- [ ] Write `IDR-005`: retire the unproven handler-order wording of `EVT-001` and replace it with the classic-observable priority-filter criterion `EVT-015`.
+- [ ] Update `CAP-001/criteria.md`: keep the `EVT-001` tombstone, mint `EVT-015`, and make the new scenario's evidence requirements explicit — serves `EVT-015`.
+- [ ] Add `EventPriorityConformanceTest.java` with positive and negative integration evidence named `EVT-015` — serves `EVT-015`.
+- [ ] Repair the bridge event-dispatch path so `EventPriorityFilter` suppresses the lower-priority scan while its wall handler is blocked — serves `EVT-015`.
+- [ ] Update the CAP-001 overview and `P-001` bookkeeping to name the retired and replacement doors — serves `EVT-015`.
+- [ ] Run the tier-1 unit suite and the focused tier-2 conformance test with the matched local Tank Royale API/runner pair; retain any environment limitation in the handoff.
+- [ ] Digest: regenerate indexes, run `clue validate`, commit the final state, and delete this change workspace.