feat(python): durable transaction subsystem and Kiro IDE 1.0 adapter

Land the crash-recoverable managed-file transaction engine (journal,
staging, commit, restart recovery, opaque snapshots, private artifacts,
Windows DACL hardening) and the Kiro IDE 1.0 workspace hook plugin with
legacy migration.

Repair five defects that kept the suite red (426 failures to zero):

- Anchor the transaction state directory at ~/.elydora/transactions
  instead of the shared system temp directory. The recovery journal now
  shares fate with the runtimes it repairs, survives temp cleaners, and
  isolates naturally under per-test fake homes.
- Skip pending-transaction recovery when no state directory exists so
  rejected installs never create runtime directories.
- Create the runtime root's parent before hardening it, matching the Go
  SDK's expand behavior for fresh homes.
- Stop using file timestamps as identity evidence: Windows exposes
  mtime/ctime with delayed cross-handle visibility. Reads now prove
  stability through a verification re-read; snapshots compare device,
  inode, size, and SHA-256.
- Stop comparing synthetic Windows permission bits across stat
  boundaries: path stats derive execute bits from file extensions while
  handle stats do not, which falsely flagged every managed .cmd wrapper.
  Snapshots store raw modes for restoration and compare portable bits.

Reservation cleanup now verifies content before discarding so a
concurrent writer's file at a rollback path is preserved, and the
install-failure tests assert the precise contract: no agent runtime
residue, with the durable state directory allowed to remain.

Gates: pytest 807 passed, mypy clean, pip check clean, wheel builds.

elydora/plugins/_opaque_cleanup.py

@@ -0,0 +1,64 @@
+"""Identity-bound cleanup for captured opaque rollback files."""
+
+from __future__ import annotations
+
+from ._opaque_removal import (
+    StagedOpaqueRemoval,
+    _same_captured,
+    stream_opaque_snapshot,
+)
+from ._private_artifact import QuarantinedFile
+
+
+def _capture_cleanup_physical(
+    staged: StagedOpaqueRemoval,
+    source_path: str,
+) -> None:
+    del staged, source_path
+
+
+def cleanup_opaque_removal(staged: StagedOpaqueRemoval) -> None:
+    if not staged.committed or staged.rollback_path is None:
+        return
+    expected = staged.rollback_snapshot
+    if expected is None:
+        raise OSError(f"Missing rollback data for {staged.change.label}")
+    rollback_path = staged.rollback_path
+    _capture_cleanup_physical(staged, rollback_path)
+    captured_file = QuarantinedFile.capture(
+        staged.directory,
+        rollback_path,
+        f"{staged.change.label} rollback cleanup",
+    )
+    captured = stream_opaque_snapshot(
+        captured_file.container,
+        captured_file.entry_path,
+        f"{staged.change.label} rollback cleanup",
+        require_owner_only=False,
+    )
+    if not _same_captured(captured, expected):
+        try:
+            captured_file.restore(rollback_path)
+        except Exception as error:
+            preserved = (
+                captured_file.preserve()
+                if captured_file.active
+                else rollback_path
+            )
+            staged.rollback_path = None
+            raise OSError(
+                f"{staged.change.label} changed at cleanup boundary; "
+                f"concurrent data preserved at {preserved}: {error}"
+            ) from error
+        staged.rollback_path = None
+        raise OSError(
+            f"{staged.change.label} changed at cleanup boundary; "
+            f"concurrent data preserved at {rollback_path}"
+        )
+    captured_file.discard()
+    staged.rollback_path = None
+    staged.rollback_snapshot = None
+    staged.committed = False
+
+
+__all__ = ["cleanup_opaque_removal"]