Merge branch 'main' into craig/display-status

CHANGELOG.md

@@ -5,6 +5,7 @@
 - Reasoning: `reasoning_tokens` for setting maximum reasoning tokens (currently only supported by Claude Sonnet 3.7)
 - Reasoning: `reasoning_history` can now be specified as "none", "all", "last", or "auto" (which yields a provider specific recommended default).
 - Web Browser: [Various improvements](https://github.com/UKGovernmentBEIS/inspect_ai/pull/1314) to performance and robustness along with several bug fixes.
+- OpenAI: Provide long connection (reasoning friendly) socket defaults in http client 
 - OpenAI: Capture `reasoning_tokens` when reported.
 - OpenAI: Retry on rate limit requests with "Request too large".
 - OpenAI: Tolerate `None` for assistant content (can happen when there is a refusal).
@@ -16,6 +17,8 @@
 - Bugfix: Catch standard `TimeoutError` for running shell commands in the computer tool container.
 - Bugfix: Fix audio and video inputs for new Google GenAI client.
 - Bugfix: Ensure that token limits are not enforced during model graded scoring.
+- Inspect View: Fix layout issues with human agent terminal session playback.
+- Inspect View: Improve tool input / output appearance when rendered in VSCode.
 
 ## v0.3.70 (25 February 2025)
 

src/inspect_ai/_view/www/App.css

@@ -805,15 +805,19 @@ table.table.table-sm td {
   overflow: unset;
 }
 
+.markdown-content pre[class*="language-"],
 pre[class*="language-"].tool-output,
 .tool-output {
   background-color: #f8f8f8;
 }
+
+.vscode-dark .markdown-content pre[class*="language-"],
 .vscode-dark pre[class*="language-"].tool-output,
 .vscode-dark .tool-output {
   background-color: #333333;
 }
 
+.markdown-content pre[class*="language-"],
 pre[class*="language-"].tool-output {
   border: none !important;
   box-shadow: none !important;