codex-as-api 0.8.1__tar.gz → 0.8.2__tar.gz

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: codex-as-api
-Version: 0.8.1
+Version: 0.8.2
 Summary: Use ChatGPT/Codex OAuth as a local OpenAI-compatible API server.
 Project-URL: Homepage, https://github.com/Eunho-J/codex-as-api
 Project-URL: Repository, https://github.com/Eunho-J/codex-as-api

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/RELEASES.md

@@ -1,5 +1,21 @@
 # Release Notes
 
+## v0.8.2
+
+### Hotfix: disjoint Anthropic cache usage (all runtimes)
+
+- Correct Anthropic Messages usage in streaming and non-streaming responses:
+  `input_tokens` now excludes both cache reads and reported cache creation,
+  which remain in their separate `cache_read_input_tokens` and
+  `cache_creation_input_tokens` counters. Previously, consumers adding those
+  counters counted cached input twice.
+- Keep OpenAI/Responses totals unchanged. Preserve unknown cache counters as
+  null; subtract only reported cache buckets. Reject impossible cache sums
+  exceeding the upstream input total instead of emitting negative/clamped usage.
+- Add no-cache, partial/full-cache, cache-write, invalid-total and HTTP pipeline
+  regressions across TypeScript, Python and Rust. Example: 100 total input,
+  50 cache read and 25 cache write becomes 25 uncached + 50 read + 25 write.
+
 ## v0.8.1
 
 ### Explicit Ultra Fast selection (all runtimes)

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "codex-as-api"
-version = "0.8.1"
+version = "0.8.2"
 description = "Use ChatGPT/Codex OAuth as a local OpenAI-compatible API server."
 readme = "README.md"
 license = { text = "Apache-2.0" }

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/src/codex_as_api/__init__.py

@@ -1,3 +1,3 @@
 """codex-as-api: Use ChatGPT/Codex OAuth as a local OpenAI-compatible API server."""
 
-__version__ = "0.8.1"
+__version__ = "0.8.2"

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/src/codex_as_api/anthropic_adapter.py

@@ -718,13 +718,23 @@ def _reject_unrepresentable_response_events(raw: dict | None) -> None:
         )
 
 
+def _uncached_input_tokens(total: int, cache_read: int | None, cache_write: int | None) -> int:
+    # OpenAI totals include cache buckets; Anthropic exposes disjoint counters.
+    cached = (cache_read or 0) + (cache_write or 0)
+    if cached > total:
+        raise ChatGPTOAuthProtocolError("usage cache tokens must not exceed input_tokens")
+    return total - cached
+
+
 def _anthropic_usage_from_internal(usage: Usage) -> dict[str, Any]:
     return {
         "cache_creation": None,
         "cache_creation_input_tokens": usage.cache_write_tokens,
         "cache_read_input_tokens": usage.cached_tokens,
         "inference_geo": None,
-        "input_tokens": usage.prompt_tokens,
+        "input_tokens": _uncached_input_tokens(
+            usage.prompt_tokens, usage.cached_tokens, usage.cache_write_tokens
+        ),
         "iterations": None,
         "output_tokens": usage.completion_tokens,
         "server_tool_use": None,
@@ -910,7 +920,7 @@ def _anthropic_usage_from_provider(usage: Any) -> dict[str, Any]:
     out: dict[str, Any] = {
         "cache_creation_input_tokens": cache_write,
         "cache_read_input_tokens": cache_read,
-        "input_tokens": input_tokens,
+        "input_tokens": _uncached_input_tokens(input_tokens, cache_read, cache_write),
         "iterations": None,
         "output_tokens": output_tokens,
         "server_tool_use": server_tool_use,

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/src/codex_as_api/server.py

@@ -497,7 +497,7 @@ def _health_error_message(exc: BaseException) -> str:
 app = FastAPI(
     title="codex-as-api",
     description="Local OpenAI-compatible API server backed by ChatGPT/Codex OAuth.",
-    version="0.8.1",
+    version="0.8.2",
 )
 
 

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/tests/test_anthropic_adapter.py

@@ -1206,6 +1206,23 @@ class TestInternalResponseToAnthropic:
         result = internal_response_to_anthropic(resp, "m", "msg_1")
         assert result["usage"]["cache_read_input_tokens"] == 50
         assert result["usage"]["cache_creation_input_tokens"] == 25
+        assert result["usage"]["input_tokens"] == 25
+
+    @pytest.mark.parametrize("read,write", [(0, 0), (40, 0), (100, 0), (0, 100), (40, 60), (101, 0), (60, 41)])
+    def test_disjoint_cache_buckets(self, read, write):
+        source = Usage(prompt_tokens=100, completion_tokens=7, total_tokens=107,
+                       cached_tokens=read, cache_write_tokens=write)
+        response = AssistantResponse(content="hi", tool_calls=(), finish_reason="stop",
+                                     usage=source, reasoning_content=None, raw=None)
+        if read + write > 100:
+            with pytest.raises(ChatGPTOAuthProtocolError, match="cache tokens must not exceed"):
+                internal_response_to_anthropic(response, "m", "msg")
+        else:
+            usage = internal_response_to_anthropic(response, "m", "msg")["usage"]
+            assert usage["input_tokens"] == 100 - read - write
+            buckets = ("input_tokens", "cache_read_input_tokens", "cache_creation_input_tokens")
+            assert sum(usage[k] for k in buckets) == 100
+            assert source.prompt_tokens == 100
 
     def test_web_search_provider_output_is_rejected(self):
         resp = AssistantResponse(
@@ -1460,7 +1477,7 @@ class TestAnthropicStreamAdapter:
         assert msg_delta["usage"] == {
             "cache_creation_input_tokens": 11,
             "cache_read_input_tokens": 13,
-            "input_tokens": 123,
+            "input_tokens": 99,
             "iterations": None,
             "output_tokens": 7,
             "server_tool_use": {"web_search_requests": 2, "web_fetch_requests": 1},
@@ -1516,6 +1533,22 @@ class TestAnthropicStreamAdapter:
         with pytest.raises(ChatGPTOAuthProtocolError, match="provider (finish )?usage"):
             self._collect_events(events)
 
+    @pytest.mark.parametrize("read,write", [(0, 0), (40, 0), (100, 0), (0, 100), (40, 60), (101, 0), (60, 41)])
+    def test_stream_disjoint_cache_buckets(self, read, write):
+        source = {"input_tokens": 100, "output_tokens": 7, "total_tokens": 107,
+                  "input_tokens_details": {"cached_tokens": read, "cache_write_tokens": write}}
+        events = [{"type": "finish", "finish_reason": "stop", "usage": source}]
+        if read + write > 100:
+            with pytest.raises(ChatGPTOAuthProtocolError, match="cache tokens must not exceed"):
+                self._collect_events(events)
+        else:
+            result = self._collect_events(events)
+            usage = next(e["usage"] for e in result if e["type"] == "message_delta")
+            assert usage["input_tokens"] == 100 - read - write
+            buckets = ("input_tokens", "cache_read_input_tokens", "cache_creation_input_tokens")
+            assert sum(usage[k] for k in buckets) == 100
+            assert source["input_tokens"] == 100
+
     def test_maps_responses_cache_write_details_into_message_delta(self):
         result = self._collect_events(
             [

{codex_as_api-0.8.1 → codex_as_api-0.8.2}/tests/test_provider.py

@@ -2158,7 +2158,7 @@ def test_codex_cli_headers_include_official_originator_and_versioned_user_agent(
 
     assert headers["originator"] == "codex_cli_rs"
     assert headers["User-Agent"].startswith("codex_cli_rs/0.153.3 (")
-    assert headers["User-Agent"].endswith(") codex-as-api/0.8.1")
+    assert headers["User-Agent"].endswith(") codex-as-api/0.8.2")
 
 
 def test_codex_cli_headers_reject_unpinned_version():
@@ -2881,7 +2881,7 @@ def test_provider_headers_include_codex_cli_headers(auth_json_factory):
 
     assert headers["originator"] == "codex_cli_rs"
     assert headers["User-Agent"].startswith("codex_cli_rs/0.153.3 (")
-    assert headers["User-Agent"].endswith(") codex-as-api/0.8.1")
+    assert headers["User-Agent"].endswith(") codex-as-api/0.8.2")
     assert headers["Authorization"].startswith("Bearer ")