chore: sync web dist from code-app

[wbxl2000]

Related Issue

N/A — routine web bundle sync from code-app (internal workflow).

Problem

Ship the latest web UI from code-app (054a6e98) in the CLI's prebuilt bundle apps/kimi-code/dist-web.

What changed

Synced apps/kimi-code/dist-web from code-app at 054a6e98b9c992ee4fe5f83c0091b7e931561aba, with the corresponding changesets. User-facing web changes in this sync include:

• Tower multi-agent collaboration mode (experimental), via /tower or the composer plus menu
• Selection quoting / comments: restored and revamped in messages, file previews, diff and per-turn changes panels, and the terminal
• Per-turn file change stats fixes (accurate added/removed line counts after repeated edits)
• Permission modes renamed to Always Ask / Ask When Needed / Never Ask
• Fixes: default thinking effort not settable to Max, background agent cards showing a running indicator, Esc no longer closing the right detail panel, quote pill styling, tooltip hover delay

Checklist

• I have read the CONTRIBUTING document.
• I have linked a related issue (external PRs: the issue must have a maintainer's /approve). — N/A, internal sync
• I have added tests that prove my feature works. — N/A, prebuilt bundle sync; packaging is guarded by apps/kimi-code/scripts/check-web-assets.mjs
• Ran gen-changesets skill, or this PR needs no changeset. — changesets included in the synced commit
• Ran gen-docs skill, or this PR needs no doc update. — no CLI doc update needed

[changeset-bot]

🦋 Changeset detected

Latest commit: a5311f4

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 1 package

Name	Type
@moonshot-ai/kimi-code	Minor

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[chatgpt-codex-connector]

💡 Codex Review

https://github.com/MoonshotAI/kimi-code/blob/a5311f47e9afd92f1877cb3a8b50927dcc7d953b/apps/kimi-code/dist-web/assets/index-CiHMlsuo.js#L751
Preserve existing modes when tower activation fails

When /tower <base> is invoked while plan or swarm mode is active, this request sends planMode: false and swarmMode: false together with the tower activation. applySessionAgentConfig() processes those fields first and only afterwards calls tower.enter() (packages/kap-server/src/routes/sessionAgentConfig.ts:42-66), so an invalid base, disabled experiment, or conflicting workspace owner leaves the previous mode permanently disabled even though tower activation failed. Disable the prior mode only after successful entry, or restore it on failure in the source UI before resyncing this bundle.

AGENTS.md reference: AGENTS.md:L18-L18

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".