anthropic observability

contents/docs/ai-engineering/_snippets/anthropic.mdx

@@ -0,0 +1,49 @@
+Start by installing the Anthropic Python SDK:
+
+```bash
+pip install anthropic
+```
+
+In the spot where you initialize the Anthropic SDK, import PostHog and our Anthropic wrapper, initialize PostHog with your project API key and host (from [your project settings](https://us.posthog.com/settings/project)), and pass it to our Anthropic wrapper.
+
+```python
+from posthog.ai.anthropic import Anthropic
+import posthog
+
+posthog.project_api_key = "<ph_project_api_key>"
+posthog.host = "<ph_client_api_host>"
+
+client = Anthropic(
+    api_key="sk-ant-api...",  # Replace with your Anthropic API key
+    posthog_client=posthog
+)
+``` 
+
+> **Note:** This also works with the `AsyncAnthropic` client as well as `AnthropicBedrock`, `AnthropicVertex`, and the async versions of those.
+
+Now, when you use the Anthropic SDK, it automatically captures many properties into PostHog including `$ai_input`, `$ai_input_tokens`, `$ai_latency`, `$ai_model`, `$ai_model_parameters`, `$ai_output_choices`, and `$ai_output_tokens`.
+
+You can also capture additional properties like `posthog_distinct_id`, `posthog_trace_id`, `posthog_properties`, `posthog_groups`, and `posthog_privacy_mode`.
+
+```python
+response = client.messages.create(
+    model="claude-3-opus-20240229",
+    messages=[
+        {
+            "role": "user",
+            "content": "Tell me a fun fact about hedgehogs"
+        }
+    ],
+    posthog_distinct_id="user_123", # optional
+    posthog_trace_id="trace_123", # optional
+    posthog_properties={"conversation_id": "abc123", "paid": True}, # optional
+    posthog_groups={"company": "company_id_in_your_db"},  # optional 
+    posthog_privacy_mode=False # optional
+)
+
+print(response.content[0].text)
+```
+
+> **Notes:** 
+> - This also works when message streams are used (e.g. `stream=True` or `client.messages.stream(...)`).
+> - If you want to capture LLM events anonymously, **don't** pass a distinct ID to the request. See our docs on [anonymous vs identified events](/docs/data/anonymous-vs-identified-events) to learn more. 

contents/docs/ai-engineering/_snippets/openai.mdx

@@ -21,7 +21,7 @@ client = OpenAI(
 
 > **Note:** This also works with the `AsyncOpenAI` client. 
 
-Now, when you use the OpenAI SDK, it automatically captures many properties into PostHog including `$ai_input`, `$ai_input_tokens`, `$ai_latency`, `$ai_model`, `$ai_model_parameters`, `$ai_output`, and `$ai_output_tokens`.
+Now, when you use the OpenAI SDK, it automatically captures many properties into PostHog including `$ai_input`, `$ai_input_tokens`, `$ai_latency`, `$ai_model`, `$ai_model_parameters`, `$ai_output_choices`, and `$ai_output_tokens`.
 
 You can also capture additional properties like `posthog_distinct_id`, `posthog_trace_id`, `posthog_properties`, `posthog_groups`, and `posthog_privacy_mode`.
 
@@ -33,8 +33,8 @@ response = client.chat.completions.create(
     ],
     posthog_distinct_id="user_123", # optional
     posthog_trace_id="trace_123", # optional
-    posthog_properties={"conversation_id": "abc123", "paid": True} # optional
-    posthog_groups={"company": "company_id_in_your_db"}  # optional 
+    posthog_properties={"conversation_id": "abc123", "paid": True}, # optional
+    posthog_groups={"company": "company_id_in_your_db"},  # optional 
     posthog_privacy_mode=False # optional
 )
 

contents/docs/ai-engineering/observability.mdx

@@ -32,18 +32,23 @@ The rest of the setup depends on the LLM platform you're using. These SDKs _do n
 
 import Tab from "components/Tab"
 import OpenAIInstall from "./_snippets/openai.mdx"
+import AnthropicInstall from "./_snippets/anthropic.mdx"
 import LangChainInstall from "./_snippets/langchain.mdx"
 
 <!-- prettier-ignore -->
-<Tab.Group tabs={['OpenAI', 'Langchain']}>
+<Tab.Group tabs={['OpenAI', 'Anthropic', 'Langchain']}>
     <Tab.List>
         <Tab>OpenAI</Tab>
+        <Tab>Anthropic</Tab>
         <Tab>LangChain</Tab>
     </Tab.List>
     <Tab.Panels>
         <Tab.Panel>
             <OpenAIInstall />
         </Tab.Panel>
+        <Tab.Panel>
+            <AnthropicInstall />
+        </Tab.Panel>
         <Tab.Panel>
             <LangChainInstall />
         </Tab.Panel>
@@ -52,7 +57,7 @@ import LangChainInstall from "./_snippets/langchain.mdx"
 
 ## Privacy mode
 
-To avoid storing potentially sensitive prompt and completion data, you can enable privacy mode. This excludes the `$ai_input` and `$ai_output` properties from being captured.
+To avoid storing potentially sensitive prompt and completion data, you can enable privacy mode. This excludes the `$ai_input` and `$ai_output_choices` properties from being captured.
 
 This can be done either by setting the `privacy_mode` config option in the Python SDK like this:
 
@@ -75,6 +80,14 @@ client.chat.completions.create(
 )
 ```
 
+```python file=Anthropic
+response = client.messages.create(
+    model="claude-3-opus-20240229",
+    messages=[...],
+    posthog_privacy_mode=True
+)
+```
+
 ```python file=LangChain
 callback_handler = PosthogCallbackHandler(
     client,