Conversation
There was a problem hiding this comment.
Pull request overview
This PR broadens xAI (Grok) model detection from the specific grok-code* family to all grok* families and uses that shared classification to adjust model capabilities, prompts, and Responses API request shaping in the Copilot extension.
Changes:
- Introduces
isXAIFamily()(grok*) and uses it to drive replace-string edit-tool capability decisions. - Disables Responses API
reasoning.summaryandreasoning.effortconfiguration for xAI models. - Updates agent prompt selection / CLI prompt tagging and disables the todo-list tool for xAI models.
Show a summary per file
| File | Description |
|---|---|
| extensions/copilot/src/platform/endpoint/node/responsesApi.ts | Uses new capability helpers to gate reasoning summary/effort in Responses API request bodies (notably for xAI). |
| extensions/copilot/src/platform/endpoint/common/chatModelCapabilities.ts | Adds isXAIFamily() and new reasoning capability helpers; updates replace-string capability checks to use xAI family detection. |
| extensions/copilot/src/extension/prompts/node/agent/xAIPrompts.tsx | Broadens prompt resolver matching from grok-code* to grok* and renames the default Grok prompt class. |
| extensions/copilot/src/extension/prompts/node/agent/copilotCLIPrompt.tsx | Switches Grok detection to isXAIFamily() for deciding whether to emit the user_query tag. |
| extensions/copilot/src/extension/intents/node/agentIntent.ts | Switches xAI detection to isXAIFamily() to disable the todo-list tool for Grok models. |
Copilot's findings
- Files reviewed: 5/5 changed files
- Comments generated: 1
Screenshot ChangesBase: Changed (10) |
|
⏳ Queued vscode build for
|
|
~requires-eval-assessment terminalbench2 grok-4.3 |
|
⏳ Queued vscode build for
|
|
~requires-eval-assessment vscbench grok-4.3 |
|
⏳ Queued vscode build for
|
|
~requires-eval-assessment swebench.litmus grok-4.3 |
|
⏳ Queued vscode build for
|
|
🚀 Queued eval-assessment publish build for
|
|
🔬 Queued eval-assessment benchmark for
Results will be posted back here when the run completes. |
|
✅ Eval-assessment build published.
|
No description provided.