chore(deps): anthropic 1.5, openai 3.13, and the move to httpx2

Both SDKs' 1.x/3.x majors run on httpx2 (the maintained fork of httpx by its
original author, published by Pydantic at github.com/pydantic/httpx2, version
line 2.x) instead of httpx, which is what makes these two bumps one change.
httpx2 is now a declared dependency because two modules import it directly.
Everything else in the 0.x -> 1.x / 2.x -> 3.x change lists is absent here:
no Text Completions, no `with_raw_response`, no raw `output_format` dicts, no
Bedrock client, and `requires-python` is already 3.12.

Three code changes, all forced by the bump:

The BYOM DNS-pinning client in `docsgpt/security/safe_url.py` is handed to
`OpenAI(http_client=...)`, and the SDK rejects an old-httpx client at
construction — which would have taken the SSRF guard offline. It is built on
httpx2 now, and its `sni_hostname` extension carries a `str` rather than
ascii bytes: httpcore passes the value straight to
`ssl.SSLContext.wrap_socket`, and the truststore backend httpx2 uses for the
default system trust store encodes it instead of accepting bytes. Bytes
therefore failed every real handshake while passing the existing tests, which
stub the transport out; the test now pins the type and says why.

The stream-retry error tuple in `docsgpt/llm/base.py` named `httpx`
exceptions only. The two libraries' exception classes are unrelated types, so
after the bump the retry silently stopped firing for openai and anthropic
while still working for google-genai and elevenlabs. It now covers both
stacks, with a parametrized test over each.

anthropic 1.x dropped temperature/top_p/top_k from `messages.create`'s
signature (passing one raises TypeError) without dropping them from the API,
so the provider forwards them through `extra_body`. The wire request is
unchanged and a model that rejects them 400s exactly as before.

docsgpt/llm/base.py

@@ -3,6 +3,7 @@ from abc import ABC, abstractmethod
 from typing import ClassVar, Dict, Optional, Tuple
 
 import httpx
+import httpx2
 import openai
 
 from docsgpt.cache import gen_cache, stream_cache
@@ -20,17 +21,26 @@ logger = logging.getLogger(__name__)
 # Excludes API-level errors (4xx / 5xx status) which are not transport
 # retries — RateLimitError needs backoff, BadRequestError won't get any
 # better on retry, and the existing fallback handles both.
-_STREAM_RETRYABLE_TRANSPORT_ERRORS = (
-    httpx.RemoteProtocolError,
-    httpx.ReadError,
-    httpx.ReadTimeout,
-    httpx.WriteError,
-    httpx.WriteTimeout,
-    httpx.ConnectError,
-    httpx.ConnectTimeout,
-    httpx.PoolTimeout,
-    openai.APIConnectionError,
+# Both HTTP stacks are listed because the providers are split across them and
+# the two libraries' exception classes are unrelated types: openai and
+# anthropic run on httpx2, while google-genai, elevenlabs, qdrant-client and
+# the MCP client are still on httpx. Naming only one silently stops matching
+# for half the providers — the retry just never fires.
+_TRANSPORT_ERROR_NAMES = (
+    "RemoteProtocolError",
+    "ReadError",
+    "ReadTimeout",
+    "WriteError",
+    "WriteTimeout",
+    "ConnectError",
+    "ConnectTimeout",
+    "PoolTimeout",
 )
+_STREAM_RETRYABLE_TRANSPORT_ERRORS = tuple(
+    getattr(module, name)
+    for module in (httpx2, httpx)
+    for name in _TRANSPORT_ERROR_NAMES
+) + (openai.APIConnectionError,)
 
 
 def optional_int(value) -> Optional[int]: