release: v0.51.512 — interrupted-turn user prompt no longer replays (#4393, #4283)

Maintainer fix on top: pass-4 recovered-user keep now uses actual kept
neighbours (was forward-scan-only → adjacent-user 400). + 3 regression tests.
Codex SAFE + Opus 'ship it' + suite 9575.

Co-authored-by: kaishi00 <kaishi00@users.noreply.github.com>

Closes #4393
Closes #4283

CHANGELOG.md

@@ -3,6 +3,12 @@
 
 ## [Unreleased]
 
+## [v0.51.512] — 2026-06-19 — Release RW (interrupted-turn user prompt no longer replays)
+
+### Fixed
+
+- **A user prompt from an interrupted turn no longer replays in every subsequent request (#4283).** When a turn was interrupted, its recovered user message could be re-sent on later requests (and, in some orphaned-tool shapes, produce two adjacent same-role messages that strict providers reject with a 400). The API-message sanitizer now decides whether to keep a recovered user message based on the actual post-sanitization neighbours — it is kept only when it genuinely separates two assistant turns, and dropped otherwise — with the same logic applied in both the sanitizer and the safe-position mirror. Thanks @kaishi00.
+
 ## [v0.51.511] — 2026-06-19 — Release RV (virtual transcript renders during programmatic scrolls)
 
 ### Fixed

api/streaming.py

@@ -3448,7 +3448,7 @@ def _sanitize_messages_for_api(messages, *, cfg: dict = None):
 
     # Second pass: build the sanitized list, dropping orphaned tool messages.
     clean = []
-    for idx, msg in enumerate(messages):
+    for msg in messages:
         if not isinstance(msg, dict):
             continue
         # Skip display-only Thinking entries. They are visible transcript