fix(stream+output-cap): guard empty streams and parse OpenRouter output-cap errors (#40589)

Two isolated reliability fixes:
- chat_completion_helpers: raise on a zero-chunk stream (no finish_reason,
  no content/reasoning/tool_calls) so retry handles it instead of
  fabricating a successful empty turn.
- model_metadata: parse the OpenRouter/Nous output-cap error phrasing
  ("maximum context length is N ... (A of text input, B of tool input,
  C in the output)") so parse_available_output_tokens_from_error returns
  a real cap and the caller stops looping on it.

Salvaged from #40405 (@ashishpatel26) — took the two stream/error-parsing
fixes. The PR also bundled compression-state changes (on_session_start
clearing _previous_summary; cron session-id prefix preservation, #38788);
those touch the compression hot path and are split out for separate review.

Co-authored-by: ashishpatel26 <ashishpatel26@users.noreply.github.com>

tests/test_output_cap_parsing.py

@@ -0,0 +1,27 @@
+import pytest
+from agent.model_metadata import parse_available_output_tokens_from_error
+
+
+class TestParseOpenRouterOutputCap:
+    """OpenRouter/Nous phrase the output-cap error as a context breakdown."""
+
+    def test_openrouter_breakdown_format(self):
+        msg = ("This endpoint's maximum context length is 200000 tokens. "
+               "However, you requested about 195000 tokens "
+               "(150000 of text input, 40000 of tool input, 5000 in the output).")
+        # available output = 200000 - 150000 - 40000 = 10000
+        assert parse_available_output_tokens_from_error(msg) == 10000
+
+    def test_anthropic_format_still_works(self):
+        msg = ("max_tokens: 32768 > context_window: 200000 - "
+               "input_tokens: 190000 = available_tokens: 10000")
+        assert parse_available_output_tokens_from_error(msg) == 10000
+
+    def test_non_output_cap_error_returns_none(self):
+        assert parse_available_output_tokens_from_error("some unrelated 400 error") is None
+
+    def test_breakdown_with_no_room_returns_none(self):
+        # ctx - text - tool <= 0 -> None (don't return a non-positive cap)
+        msg = ("maximum context length is 1000 tokens "
+               "(900 of text input, 200 of tool input, 0 in the output)")
+        assert parse_available_output_tokens_from_error(msg) is None