[Afloat16]

What & why

Replay returned success before its destructor drained pending weight reads. Late failures were discarded and the session remained reusable despite failed weights.

Drain explicitly before publishing success. Propagate late errors and invalidate the session; keep destructor cleanup for earlier failure and drain once, preserving an original forward error when both fail.

Area

Engine, CLI or server (edge0-infer/).

How was it tested?

moe_test_app_generation

The target and existing generation/callback/threaded-release coverage pass. Regressions cover late tail-read errors, refusal of subsequent generation, one drain and original-error preservation; main fails the first case. The local tmpfile() test shim redirects to a workspace directory because /tmp is unavailable; it still uses a real file.

All 14 L0 checks passed in a local integration with the other focused fixes. git diff --check passed.

Local C++ environment: Linux x86_64, GCC 14.2, CPU preset. Targeted builds downgrade pre-existing ignored-attributes, dangling-reference, range-loop-construct and misleading-indentation warnings (plus maybe-uninitialized in the existing diagnostics source); these accommodations are not source changes. The all-target build is blocked by existing assets_test.cpp use of std::count_if without . A clean full L0/L1 result, GPU tests and model-asset execution are not claimed.

Checklist

• Full ctest --preset cpu -L "L0|L1" passes locally
• Added/updated focused regressions and ran the affected target
• No public interface/configuration change requiring additional documentation