fix(context): clamp -1 post-compression sentinel in sibling status paths

Whole-bug-class follow-up to the tui_gateway fix: the same -1
last_prompt_tokens sentinel (parked by conversation_compression after a
compression) leaked into other status readers, producing a raw -1 or a
NEGATIVE usage_percent on the transitional turn:

- agent/context_engine.py get_status() (the ABC default every external
  context engine inherits) — highest blast radius
- gateway/slash_commands.py /usage context line
- cli.py session usage printout

All clamped to >=0, mirroring cli.py _get_status_bar_snapshot and the
tui_gateway fix. Adds an ABC get_status sentinel-clamp regression test.

tests/agent/test_context_engine.py

@@ -120,6 +120,16 @@ class TestDefaults:
         assert status["threshold_tokens"] == 100000
         assert 0 < status["usage_percent"] <= 100
 
+    def test_default_get_status_clamps_post_compression_sentinel(self):
+        """After a compression, last_prompt_tokens is the -1 sentinel. get_status
+        must clamp it to 0 rather than export a raw -1 or a negative
+        usage_percent on the transitional turn."""
+        engine = StubEngine()
+        engine.last_prompt_tokens = -1
+        status = engine.get_status()
+        assert status["last_prompt_tokens"] == 0
+        assert status["usage_percent"] >= 0
+
     def test_on_session_reset(self):
         engine = StubEngine()
         engine.last_prompt_tokens = 999