[pull] main from Alishahryar1:main

README.md

@@ -213,6 +213,7 @@ For another API endpoint, open **Providers → Custom providers → Add provider
 | [OpenAI API](https://platform.openai.com/api-keys) | `OPENAI_API_KEY` | `openai_api/gpt-5.6-sol` |
 | [GitHub Copilot](https://docs.github.com/en/copilot/how-tos/copilot-sdk/auth/authenticate) | Connect GitHub Copilot in the Admin UI | `github_copilot/<model-id>` |
 | [xAI (Grok)](https://console.x.ai/team/default/api-keys) | `XAI_API_KEY` | `xai/grok-4.5` |
+| [Alibaba Cloud](https://www.alibabacloud.com/help/en/model-studio/get-api-key) | `ALIBABA_CLOUD_API_KEY` | `alibaba_cloud/<model-id>` |
 | [QwenCloud Token Plan](https://home.qwencloud.com/api-keys) | `QWENCLOUD_API_KEY` | `qwencloud/qwen3.7-plus` |
 | [QwenCloud Coding Plan](https://home.qwencloud.com/api-keys) | `QWENCLOUD_CODING_API_KEY` | `qwencloud_coding/qwen3.7-plus` |
 | [Together AI](https://api.together.ai/settings/api-keys) | `TOGETHER_API_KEY` | `together/zai-org/GLM-5.2` |
@@ -288,6 +289,10 @@ For another API endpoint, open **Providers → Custom providers → Add provider
 - Kimi Code subscription keys use `kimi_code/`. Kimi API credit keys use
   `kimi/`. Kimi Code plans are for personal interactive coding-agent use under
   [Kimi's community guidelines](https://www.kimi.com/code/docs/en/kimi-code/community-guidelines.html).
+- Alibaba Cloud uses a Model Studio pay-as-you-go key and defaults to Singapore.
+  For another region or workspace, set `ALIBABA_CLOUD_BASE_URL` to its
+  [OpenAI-compatible endpoint](https://www.alibabacloud.com/help/en/model-studio/base-url)
+  ending in `/compatible-mode/v1`. The key and endpoint must use the same region.
 - QwenCloud Coding Plan keys use `qwencloud_coding/`. QwenCloud Token Plan keys
   use `qwencloud/`. The keys and endpoints are not interchangeable. Coding Plan
   is for local, personal, interactive coding-agent use under the

e2e/test_admin_apply.py

@@ -190,7 +190,8 @@ def test_apply_network_error_unlocks_form_and_keeps_edits(
 
 
 def test_restart_waits_for_a_new_running_server(page: Page, admin_base_url: str):
-    pending: list[Route] = []
+    status = {"status": "running", "instance_id": "old-server"}
+    status_url = f"{admin_base_url}/admin/api/status"
     page.route(
         "**/admin/api/config/apply",
         lambda route: route.fulfill(
@@ -210,21 +211,25 @@ def test_restart_waits_for_a_new_running_server(page: Page, admin_base_url: str)
     page.goto(f"{admin_base_url}/admin")
     expect(page.locator("#messageArea")).to_have_text("")
     page.wait_for_function("!state.startupRequest && !state.startupTimer")
-    page.route("**/admin/api/status", lambda route: pending.append(route))
+    page.route(status_url, lambda route: route.fulfill(json=status))
     open_provider(page, "nvidia_nim")
     page.locator("#field-NVIDIA_NIM_API_KEY").fill("new-key")
     with page.expect_request("**/admin/api/status"):
         page.get_by_role("button", name="Save", exact=True).click()
     expect(page.locator("#applyButton")).to_have_text("Reconnecting…")
-    for status in (
+    for next_status in (
         {"status": "running", "instance_id": "old-server"},
         {"status": "stopping", "instance_id": "new-server"},
     ):
-        with page.expect_request("**/admin/api/status"):
-            pending.pop(0).fulfill(json=status)
+        with page.expect_response(
+            lambda response, expected=next_status: (
+                response.url == status_url and response.json() == expected
+            )
+        ):
+            status = next_status
         expect(page.locator("#view-providers")).to_have_attribute("inert", "")
         expect(page.locator("#dirtyState")).to_have_text("Changes saved")
-    pending.pop(0).fulfill(json={"status": "running", "instance_id": "new-server"})
+    status = {"status": "running", "instance_id": "new-server"}
     expect(page.locator('[data-provider="nvidia_nim"]')).to_be_enabled()
     expect(page.locator("#dirtyState")).to_have_text("No changes")
     expect(page.locator("#messageArea")).to_have_text("Applied")

pyproject.toml

@@ -39,6 +39,7 @@ dependencies = [
     "grpcio-tools>=1.81.1",
     "nvidia-riva-client>=2.26.0",
     "pyperclip>=1.11.0",
+    "sqlalchemy>=2.1.1",
 ]
 
 [project.scripts]

smoke/lib/config.py

@@ -75,6 +75,7 @@
     "groq": "groq/openai/gpt-oss-20b",
     "cline_pass": "cline_pass/cline-pass/deepseek-v4-flash",
     "xai": "xai/grok-4.5",
+    "alibaba_cloud": "alibaba_cloud/qwen3-coder-plus",
     "qwencloud": "qwencloud/qwen3.7-plus",
     "qwencloud_coding": "qwencloud_coding/qwen3.7-plus",
     "together": "together/zai-org/GLM-5.2",

smoke/product/test_client_product_live.py

@@ -47,6 +47,21 @@ def _json_object_lines(text: str) -> list[JsonObject]:
     return objects
 
 
+def _trace_log_events(text: str) -> list[JsonObject]:
+    events: list[JsonObject] = []
+    for row in _json_object_lines(text):
+        record = row.get("record")
+        if not isinstance(record, dict):
+            continue
+        extra = record.get("extra")
+        if not isinstance(extra, dict):
+            continue
+        payload = extra.get("trace_payload")
+        if isinstance(payload, dict):
+            events.append(payload)
+    return events
+
+
 @pytest.mark.smoke_target("clients")
 def test_vscode_protocol_e2e(smoke_config: SmokeConfig) -> None:
     provider_model = ProviderMatrixDriver(smoke_config).first_model()
@@ -985,7 +1000,7 @@ def test_claude_cli_web_search_e2e(smoke_config: SmokeConfig, tmp_path: Path) ->
     assert '"type": "server_tool_use"' in automatic_payload
     assert '"type": "web_search_tool_result"' in automatic_payload
     assert "github.com" in automatic_payload
-    log_rows = _json_object_lines(server_log)
+    log_rows = _trace_log_events(server_log)
     assert (
         sum(
             row.get("event") == "free_claude_code.api.web_search.automatic_recognized"
@@ -1087,7 +1102,7 @@ def test_claude_auto_mode_openai_connected_e2e(
     ):
         assert unexpected not in combined_lower
 
-    log_rows = _json_object_lines(server_log)
+    log_rows = _trace_log_events(server_log)
     policy_rows = [
         row
         for row in log_rows

src/free_claude_code/cli/commands.py

@@ -259,7 +259,7 @@ def _run_bound(
 
         from free_claude_code.runtime.bootstrap import build_asgi_app
 
-        from .uvicorn_server import RuntimeServer
+        from .uvicorn_server import RuntimeServer, uvicorn_log_config
 
         asgi_app = build_asgi_app(
             settings,
@@ -270,9 +270,7 @@ def _run_bound(
             host=settings.host,
             port=settings.port,
             log_level="debug",
-            log_config=(
-                uvicorn.config.LOGGING_CONFIG if self._console_logging else None
-            ),
+            log_config=uvicorn_log_config(console=self._console_logging),
             timeout_graceful_shutdown=SERVER_GRACEFUL_SHUTDOWN_SECONDS,
         )
 

src/free_claude_code/cli/uvicorn_server.py

@@ -2,9 +2,25 @@
 
 import socket
 from collections.abc import Awaitable, Callable
+from copy import deepcopy
 
 import uvicorn
 
+from free_claude_code.config.logging_config import InterceptHandler
+
+
+def uvicorn_log_config(*, console: bool) -> dict[str, object]:
+    """Send Uvicorn records to FCC's file and optional standard console handlers."""
+    config = deepcopy(uvicorn.config.LOGGING_CONFIG)
+    if not console:
+        config["formatters"] = {}
+        config["handlers"] = {}
+    config["handlers"]["fcc"] = {"()": InterceptHandler}
+    for name in ("uvicorn", "uvicorn.access"):
+        handlers = config["loggers"][name]["handlers"]
+        config["loggers"][name]["handlers"] = [*handlers, "fcc"] if console else ["fcc"]
+    return config
+
 
 class RuntimeServer(uvicorn.Server):
     """Notify the runtime before Uvicorn waits for long-lived HTTP responses."""

src/free_claude_code/config/admin/provider_manifest.py

@@ -165,6 +165,19 @@ class ProviderFieldOverride(TypedDict, total=False):
             "models."
         ),
     },
+    "ALIBABA_CLOUD_API_KEY": {
+        "description": (
+            "Alibaba Cloud Model Studio pay-as-you-go API key. Use a key from "
+            "the same region as the base URL."
+        ),
+    },
+    "ALIBABA_CLOUD_BASE_URL": {
+        "description": (
+            "Optional Model Studio OpenAI-compatible base URL ending in "
+            "/compatible-mode/v1. Defaults to Singapore. Set your regional "
+            "or workspace URL when using a key from another region."
+        ),
+    },
     "QWENCLOUD_API_KEY": {
         "label": "QwenCloud Token Plan API Key",
         "description": (

src/free_claude_code/config/logging_config.py

@@ -1,14 +1,12 @@
 """Loguru-based structured logging configuration.
 
 Structured logs are written as JSON lines to a configurable path (default
-``logs/server.log``). Stdlib logging is intercepted and funneled to loguru.
-Context vars (request_id, node_id, chat_id) from contextualize() are
-included at top level for easy grep/filter.
+``~/.fcc/logs/server.log``). Stdlib logging is intercepted and funneled to loguru.
+Loguru's JSON stores exception metadata in ``record.exception`` and context in
+``record.extra``. Structured traces live in ``record.extra.trace_payload``.
 """
 
-import json
 import logging
-import re
 import threading
 from pathlib import Path
 
@@ -30,63 +28,6 @@
     "telegram.ext",
 )
 
-# Loguru ``logger.bind()`` key used by structured TRACE payloads; ``core/trace.py``
-# uses the identical string constant ``TRACE_PAYLOAD_BINDING``.
-_TRACE_PAYLOAD_BINDING = "trace_payload"
-
-# Context keys we promote to top-level JSON for traceability / grep
-_CONTEXT_KEYS = (
-    "request_id",
-    "node_id",
-    "chat_id",
-    "claude_session_id",
-    "http_method",
-    "http_path",
-)
-
-_TELEGRAM_BOT_RE = re.compile(
-    r"(https?://api\.telegram\.org/)bot([0-9]+:[A-Za-z0-9_-]+)(/?)",
-    re.IGNORECASE,
-)
-# Authorization: Bearer <token> (HTTP client / proxy debug lines)
-_AUTH_BEARER_RE = re.compile(
-    r"(\bAuthorization\s*:\s*Bearer\s+)([^\s'\"]+)",
-    re.IGNORECASE,
-)
-
-
-def _redact_sensitive_substrings(message: str) -> str:
-    """Remove obvious API tokens and secrets before JSON log line emission."""
-    text = _TELEGRAM_BOT_RE.sub(r"\1bot<redacted>\3", message)
-    return _AUTH_BEARER_RE.sub(r"\1<redacted>", text)
-
-
-def _serialize_with_context(record) -> str:
-    """Format record as JSON with context vars at top level.
-    Returns a format template; we inject _json into record for output.
-    """
-    extra = record.get("extra", {})
-    out = {
-        "time": str(record["time"]),
-        "level": record["level"].name,
-        "message": _redact_sensitive_substrings(str(record["message"])),
-        "module": record["name"],
-        "function": record["function"],
-        "line": record["line"],
-    }
-    trace_payload = extra.get(_TRACE_PAYLOAD_BINDING)
-    for key in _CONTEXT_KEYS:
-        if key in extra and extra[key] is not None:
-            out[key] = extra[key]
-    if isinstance(trace_payload, dict):
-        for tk, tv in trace_payload.items():
-            if tk in out:
-                continue
-            out[tk] = tv
-        out["trace"] = True
-    record["_json"] = json.dumps(out, default=str)
-    return "{_json}\n"
-
 
 class InterceptHandler(logging.Handler):
     """Redirect stdlib logging to loguru."""
@@ -129,7 +70,7 @@ def _add_file_sink(log_file: str | Path, level: str) -> int:
     return logger.add(
         log_path,
         level=level,
-        format=_serialize_with_context,
+        serialize=True,
         encoding="utf-8",
         mode="a",
         rotation="50 MB",
@@ -150,7 +91,7 @@ def configure_logging(
     Idempotent: skips if already configured with the same path, level, and verbosity.
     On path or level change, replaces only the file sink without truncating.
     On verbosity change alone, updates only the third-party logger levels.
-    Use force=True to reconfigure from scratch.
+    Use force=True to rebuild handlers while preserving existing log records.
 
     When ``verbose_third_party`` is false, managed noisy third-party loggers
     are capped at WARNING unless explicitly configured otherwise.
@@ -174,8 +115,6 @@ def configure_logging(
 
         logger.remove()
 
-        log_path.write_text("")
-
         _sink_id = _add_file_sink(log_path, level)
 
         intercept = InterceptHandler()

src/free_claude_code/config/provider_catalog.py

@@ -54,6 +54,7 @@
 OPENAI_CODEX_DEFAULT_BASE = "https://chatgpt.com/backend-api/codex"
 # xAI OpenAI-compatible Chat Completions API.
 XAI_DEFAULT_BASE = "https://api.x.ai/v1"
+ALIBABA_CLOUD_DEFAULT_BASE = "https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
 # QwenCloud Token Plan OpenAI-compatible Chat Completions API.
 QWENCLOUD_DEFAULT_BASE = (
     "https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1"
@@ -219,6 +220,18 @@ def configuration_attrs(self) -> tuple[str, ...]:
         default_base_url=XAI_DEFAULT_BASE,
         proxy_attr="xai_proxy",
     ),
+    "alibaba_cloud": ProviderDescriptor(
+        provider_id="alibaba_cloud",
+        display_name="Alibaba Cloud",
+        website_url="https://www.alibabacloud.com/en/product/modelstudio",
+        logo_filename="alibabacloud-color.svg",
+        credential_env="ALIBABA_CLOUD_API_KEY",
+        credential_url="https://www.alibabacloud.com/help/en/model-studio/get-api-key",
+        credential_attr="alibaba_cloud_api_key",
+        default_base_url=ALIBABA_CLOUD_DEFAULT_BASE,
+        base_url_attr="alibaba_cloud_base_url",
+        proxy_attr="alibaba_cloud_proxy",
+    ),
     "qwencloud": ProviderDescriptor(
         provider_id="qwencloud",
         display_name="QwenCloud Token Plan",

src/free_claude_code/config/settings.py

@@ -290,6 +290,14 @@ def validate_provider_references(self) -> Settings:
         default=None, validation_alias="XAI_API_KEY"
     )
 
+    # ==================== Alibaba Cloud Model Studio ====================
+    alibaba_cloud_api_key: OptionalNonEmptyString = Field(
+        default=None, validation_alias="ALIBABA_CLOUD_API_KEY"
+    )
+    alibaba_cloud_base_url: OptionalNonEmptyString = Field(
+        default=None, validation_alias="ALIBABA_CLOUD_BASE_URL"
+    )
+
     # ==================== QwenCloud Token Plan (OpenAI-compatible) ====================
     qwencloud_api_key: OptionalNonEmptyString = Field(
         default=None, validation_alias="QWENCLOUD_API_KEY"
@@ -433,6 +441,9 @@ def validate_provider_references(self) -> Settings:
     xai_proxy: OptionalNonEmptyString = Field(
         default=None, validation_alias="XAI_PROXY"
     )
+    alibaba_cloud_proxy: OptionalNonEmptyString = Field(
+        default=None, validation_alias="ALIBABA_CLOUD_PROXY"
+    )
     qwencloud_proxy: OptionalNonEmptyString = Field(
         default=None, validation_alias="QWENCLOUD_PROXY"
     )

src/free_claude_code/core/trace.py

@@ -1,6 +1,6 @@
 """Structured DEBUG traces for end-to-end request / CLI / provider logging.
 
-Emitted lines are merged into JSON log rows by ``config.logging_config``.
+Trace payloads are preserved in ``record.extra.trace_payload`` by Loguru.
 Conversation and Claude Code prompts are logged verbatim unless values live under
 sanitized credential keys (e.g. ``api_key``, ``authorization``). The default
 INFO log level excludes these detailed request traces.
@@ -50,7 +50,7 @@ def sanitize_trace_value(obj: Any) -> Any:
 
 
 def trace_event(*, stage: str, event: str, source: str, **fields: Any) -> None:
-    """Emit one structured DEBUG trace row merged into JSON by the log sink."""
+    """Emit a DEBUG record with a structured payload bound to its extra fields."""
     payload = sanitize_trace_value(
         {
             "stage": stage,

src/free_claude_code/providers/alibaba_cloud/__init__.py

@@ -0,0 +1,5 @@
+"""Alibaba Cloud provider exports."""
+
+from .client import AlibabaCloudProvider
+
+__all__ = ["AlibabaCloudProvider"]