Merge origin/main into codex/align-codex-provider-conventions-mainrepo

tools/openrouter_client.py

@@ -25,7 +25,15 @@ def get_async_client() -> AsyncOpenAI:
         api_key = os.getenv("OPENROUTER_API_KEY")
         if not api_key:
             raise ValueError("OPENROUTER_API_KEY environment variable not set")
-        _client = AsyncOpenAI(api_key=api_key, base_url=OPENROUTER_BASE_URL)
+        _client = AsyncOpenAI(
+            api_key=api_key,
+            base_url=OPENROUTER_BASE_URL,
+            default_headers={
+                "HTTP-Referer": "https://github.com/NousResearch/hermes-agent",
+                "X-OpenRouter-Title": "Hermes Agent",
+                "X-OpenRouter-Categories": "cli-agent",
+            },
+        )
     return _client
 
 

tools/session_search_tool.py

@@ -170,12 +170,15 @@ async def _summarize_session(
     max_retries = 3
     for attempt in range(max_retries):
         try:
+            from agent.auxiliary_client import get_auxiliary_extra_body
+            _extra = get_auxiliary_extra_body()
             response = await _async_aux_client.chat.completions.create(
                 model=_SUMMARIZER_MODEL,
                 messages=[
                     {"role": "system", "content": system_prompt},
                     {"role": "user", "content": user_prompt},
                 ],
+                **({} if not _extra else {"extra_body": _extra}),
                 temperature=0.1,
                 max_tokens=MAX_SUMMARY_TOKENS,
             )

tools/vision_tools.py

@@ -314,11 +314,14 @@ async def vision_analyze_tool(
         logger.info("Processing image with %s...", model)
         
         # Call the vision API
+        from agent.auxiliary_client import get_auxiliary_extra_body
+        _extra = get_auxiliary_extra_body()
         response = await _aux_async_client.chat.completions.create(
             model=model,
             messages=messages,
             temperature=0.1,
             max_tokens=2000,
+            **({} if not _extra else {"extra_body": _extra}),
         )
         
         # Extract the analysis

tools/web_tools.py

@@ -242,6 +242,8 @@ Create a markdown summary that captures all key information in a well-organized,
             if _aux_async_client is None:
                 logger.warning("No auxiliary model available for web content processing")
                 return None
+            from agent.auxiliary_client import get_auxiliary_extra_body
+            _extra = get_auxiliary_extra_body()
             response = await _aux_async_client.chat.completions.create(
                 model=model,
                 messages=[
@@ -250,6 +252,7 @@ Create a markdown summary that captures all key information in a well-organized,
                 ],
                 temperature=0.1,
                 max_tokens=max_tokens,
+                **({} if not _extra else {"extra_body": _extra}),
             )
             return response.choices[0].message.content.strip()
         except Exception as api_error:
@@ -362,6 +365,8 @@ Create a single, unified markdown summary."""
                 fallback = fallback[:max_output_size] + "\n\n[... truncated ...]"
             return fallback
 
+        from agent.auxiliary_client import get_auxiliary_extra_body
+        _extra = get_auxiliary_extra_body()
         response = await _aux_async_client.chat.completions.create(
             model=model,
             messages=[
@@ -370,6 +375,7 @@ Create a single, unified markdown summary."""
             ],
             temperature=0.1,
             max_tokens=4000,
+            **({} if not _extra else {"extra_body": _extra}),
         )
         final_summary = response.choices[0].message.content.strip()