Conversation
Remove the built-in Chat with Mux workspace, agent, docs, and the frontend/backend special-casing that supported it. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$21.45`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=21.45 -->
|
Preview deployment for your docs. Learn more about Mintlify Previews.
|
|
@codex review |
![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: 5165104d4f
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".
Preserve global mux tool scope for legacy system workspaces, keep new-chat startup away from hidden system workspaces, and stabilize the affected Bun test suites after removing Chat with Mux. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Addressed the validation regressions and the two review comments:
|
|
Codex Review: Didn't find any major issues. What shall we delve into next? 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". |
Reload menu-heavy components through fresh module instances and re-register their local mocks inside each test setup so full-suite runs don't inherit stale module state from earlier files. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Follow-up after the full-suite rerun: I tightened the SectionHeader/ProjectSidebar/AgentListItem tests so they re-register their local mocks and reload fresh component modules per test setup. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1b416a4012
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".
Map persisted `mux` agent selections to the default built-in fallback so upgraded installs keep UI/backend agent state aligned after removing the Chat with Mux agent. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Addressed the |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 4efd25bc8f
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".
Keep hidden system workspaces out of the landing-page recents list so upgraded installs do not resurface the removed Chat with Mux workspace. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Addressed the landing-page follow-up by filtering hidden system workspaces out of recent workspaces, with a focused regression test around the helper. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 467bfc39e4
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".
Preserve workspaces in the landing-page recents list while still filtering hidden legacy system workspaces from upgraded installs. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Follow-up to the landing-page recent-workspaces fix: hidden legacy system workspaces are filtered, but |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: f45851d4e9
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".
Allow visible workspaces to participate in the desktop new-chat startup fallback while still filtering hidden legacy system workspaces. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Follow-up to the recent-workspace compatibility fixes: |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6a8268229d
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".
Keep desktop startup and landing-page recents hidden when project metadata is unavailable, while still allowing visible multi-project workspaces through the explicit path. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Follow-up to the recent-workspace filtering fixes: both landing-page recents and desktop startup now fail closed when project metadata is missing, while |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6a1bc02615
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".
Defer marking desktop new-chat startup fallback as satisfied until a viable recent workspace is found, so the effect can retry after project metadata refreshes. --- _Generated with `mux` • Model: `openai:gpt-5.4` • Thinking: `xhigh` • Cost: `$49.43`_ <!-- mux-attribution: model=openai:gpt-5.4 thinking=xhigh costs=49.43 -->
|
@codex review Follow-up to the startup fallback filtering fixes: desktop new-chat now only marks the fallback as handled once it actually finds a viable recent workspace, so it can retry after project metadata refreshes. Added a retry regression test, and |
|
Codex Review: Didn't find any major issues. Can't wait for the next one! 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". |
After the Chat with Mux removal (#3123), the conditional that prepended the mux-chat workspace was deleted, leaving behind a trivial `const workspaces = inputWorkspaces;` alias. Inline the destructured name directly so the intermediate variable is gone.
2 participants
Summary
Remove the built-in Chat with Mux workspace and agent, along with the frontend, backend, docs, and test special-casing that supported it, then follow through on the compatibility and validation regressions that surfaced once the feature was gone.
Background
Chat with Mux added a hidden system workspace, a dedicated built-in agent, and multiple product-specific exceptions across routing, workspace creation, tool scope, MCP/secrets handling, docs, Storybook, and tests. Removing the feature wholesale simplifies those paths, but it also exposed a few places that still depended on legacy system-workspace behavior, persisted
muxagent ids, or test-only globals/module state that had been masked before.Implementation
_multiworkspaces project-scopedmuxagent ids to the default built-in fallback so upgraded installs keep UI/backend agent state alignedValidation
make static-checkmake testRisks
muxagent ids still need compatibility fallbacks on upgrade, so regressions there remain low-risk but worth keeping under regression coverage.Generated with
mux• Model:openai:gpt-5.4• Thinking:xhigh• Cost:$49.43