test: repair full suite — hang fix, stale-test updates, patch hygiene, timeout backstop

Suite went from 27 failures plus one permanent hang (never finished) to
1846 passed / 0 failed in ~85s, including under FORCE_COLOR.

- office_shutdown_lifecycle: construct WSHandler via the real __init__
  (helper) instead of hand-copied __new__ stubs that drift from the
  constructor (#11 added _runtime_status_sync_task and the stubs hung);
  the formerly-hanging wait now has a 5s wait_for.
- import-time patch hygiene: company_recruiter / company_reorg /
  engine_session_defaults replaced module-level permanent
  tempfile.TemporaryDirectory monkeypatching with paired
  setUpModule/tearDownModule, fixing order-dependent sqlite failures in
  transcript_pagination during full runs.
- stale tests updated to current product semantics: resume stubs use
  status="done" (failed is deliberately non-resumable), fix4 asserts the
  native review contract through build_company_work_item_contract,
  delivery fixture carries user_visible/feedback_scope=final, ownership
  doc names progress_log, session compression calls
  maybe_compact_session(force=True) explicitly, hard delete removes the
  work item row, parallel-isolation asserts delegate rebind and stubs
  _get_project_delegate, role update goes through OrgService on an
  editable custom org (plus read-only rejection case), collab_rpc patches
  the single os.name decision point instead of poisoning pathlib, codex
  no-pty builds inside the patch, identity-guard false positives reworded.
- cli_board actions rewritten against the real OfficeServiceFactory seam
  with a tempdir OPC_HOME (old direct-engine stubs were never consulted
  and the tests wrote into the real OPC home).
- cli_app assertions strip ANSI via _plain_output so a color-forcing
  shell (FORCE_COLOR) cannot break plain-text expectations.
- deleted never-runnable test_org_concurrency (pytest.mark.asyncio
  without the plugin, stdlib-only assertions) and three dead skipped
  filesystem-handoff tests.
- pyproject: dev extra (pytest, pytest-timeout) and a 300s per-test
  timeout backstop so a wedged test fails instead of stalling the suite.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

tests/test_session_context_compression.py

@@ -202,14 +202,13 @@ class SessionContextCompressionTests(unittest.IsolatedAsyncioTestCase):
             store = OPCStore(db_path)
             await store.initialize()
             memory = MemoryManager(root, "proj1", store=store)
-            memory.set_history_compactor(
-                HistoryCompactor(
-                    llm=_StubLLM(),
-                    store=store,
-                    memory_manager=memory,
-                    compression_threshold=0.85,
-                )
+            compactor = HistoryCompactor(
+                llm=_StubLLM(),
+                store=store,
+                memory_manager=memory,
+                compression_threshold=0.85,
             )
+            memory.set_history_compactor(compactor)
             session_id = "session-restart"
 
             for idx in range(6):
@@ -217,6 +216,20 @@ class SessionContextCompressionTests(unittest.IsolatedAsyncioTestCase):
                 payload = f"message {idx} " + ("alpha beta gamma delta " * 40)
                 await memory.append_session_message(session_id=session_id, role=role, text=payload)
 
+            # Append no longer auto-compacts; compaction is an explicit call.
+            # Forced compaction consumes everything up to the latest message,
+            # so the raw tail comes from messages appended afterwards.
+            compacted = await compactor.maybe_compact_session(
+                project_id="proj1",
+                session_id=session_id,
+                force=True,
+            )
+            self.assertTrue(compacted)
+            for idx in range(6, 8):
+                role = "user" if idx % 2 == 0 else "assistant"
+                payload = f"message {idx} " + ("alpha beta gamma delta " * 40)
+                await memory.append_session_message(session_id=session_id, role=role, text=payload)
+
             snapshot = await store.get_latest_session_memory_snapshot(session_id)
             self.assertIsNotNone(snapshot)
             assert snapshot is not None
@@ -225,7 +238,7 @@ class SessionContextCompressionTests(unittest.IsolatedAsyncioTestCase):
             context = await memory.build_session_prompt_context(session_id)
             self.assertIn("## Session Memory", context)
             self.assertIn("Session history summary before restart", context)
-            self.assertIn("message 5", context)
+            self.assertIn("message 7", context)
 
             await store.close()
 
@@ -236,7 +249,7 @@ class SessionContextCompressionTests(unittest.IsolatedAsyncioTestCase):
             restarted_context = await restarted_memory.build_session_prompt_context(session_id)
 
             self.assertTrue(any("Session history summary before restart" in item["content"] for item in restarted_history))
-            self.assertTrue(any("message 5" in item["content"] for item in restarted_history))
+            self.assertTrue(any("message 7" in item["content"] for item in restarted_history))
             self.assertFalse(any("message 0" in item["content"] for item in restarted_history))
             self.assertNotIn("message 0", restarted_context)