@@ -41,7 +41,7 @@ In the [Claude Console](https://platform.claude.com/), open the **OpenTelemetry*
 - **Content mode**: Select **Full content** to see prompts, responses, and tool input and output in Sentry. Select **Metadata only** to send only timing, status, and token counts.
 - **Set as organization default**: Select this for your first destination.
 
-![Add destination dialog in the Claude Console with Sentry selected as the provider](./img/claude-platform-otel-console-sentry-destination.jpg)
+![Add destination dialog in the Claude Console with Sentry selected as the provider =500x](./img/claude-platform-otel-console-sentry-destination.jpg)
 
 To send a workspace's traces to a different Sentry project, add a destination to that workspace. A workspace destination replaces the organization default for that workspace.
 
@@ -53,7 +53,7 @@ Send a Messages API request with an API key from a workspace that uses the desti
 
 Then open [**Explore > Agents**](https://sentry.io/orgredirect/organizations/:orgslug/explore/agents/) in Sentry. Your session or request shows up on the **Conversations**, **Traces**, and **LLM Calls** tabs. Select a conversation and open the **Timeline** tab to see each turn, model request, and tool call with its duration and token usage.
 
-![Sentry conversation timeline for a Claude Managed Agents session, showing model request and tool call spans](./img/claude-platform-otel-conversation-timeline.jpg)
+![Sentry conversation timeline for a Claude Managed Agents session, showing model request and tool call spans =700x](./img/claude-platform-otel-conversation-timeline.jpg)
 
 The Claude Platform puts the conversation on agent spans, and tool input and output on tool spans. To read prompts and tool results, select a tool span or open the **Transcript** tab. Model request spans don't include prompts.