Repository navigation
Keep Ranger text selection and show model reasoning - #456
Conversation
Terminal output no longer refocuses xterm while text is selected or being selected in Ranger or another overlay, so replies can be selected while agents stream. Activating a pane still reclaims focus, and Escape clears a panel selection before closing Ranger. Ranger now shows model reasoning in a collapsible Thinking section, one step per model call. It stays open while the model reasons, keeps the newest reasoning within a 16,000 character limit, omits redacted reasoning, and is shed from older turns before chat history. Also polish the panel: the model picker sizes to its name, the run status reads as a status rather than a tab, the chat subtitle shows the model label, and two tool names get readable labels. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
@codex review |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 39fffa59b0
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
When one code unit of budget remained over an older step ending in an astral character, the repaired cut was empty but still kept, rendering a blank step and undercounting skipped steps. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
@codex review |
|
Codex Review: Didn't find any major issues. 👍 Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
1 participant
Summary
list_ranger_tasks/send_user_notificationget readable labels.Snapshots gain optional
thinking,thinking_skippedandthinking_truncatedfields on assistant messages; older clients ignore them.Verification
bun run typecheck:bun bun run lint bun run build:web bun test shared/assistant.test.ts server/src/assistant/service.test.ts server/src/assistant/pi-driver.test.ts web/src/components/AssistantThinking.test.tsx web/src/components/AssistantPanel.test.ts web/src/components/AssistantTasks.test.ts web/src/terminalFocus.test.ts web/src/components/TerminalView.inputMode.test.ts bun run test:quicktest:quicktimed out two isolated-runtime DOM tests on a loaded machine (AssistantPanel.test.ts,TerminalView.inputMode.test.ts); both pass when run alone. The panel was checked manually against a dev bridge for the selector layout, status indicator, model label and tool labels. Reasoning display with a live model has not been checked manually.🤖 Generated with Claude Code