Repository navigation
chore: remove unused code and brittle tests
UI Consistency: No findings
No UI consistency violations found at ea2122a.
The head commit is unchanged from the previously reviewed revision, and the four earlier findings were answered and resolved by the author. Re-checking them against the repository's committed test policy (AGENTS.md line 107: "Do not render components to static markup to assert props or attributes, or add tests that merely assert callback wiring or mirror the implementation"), the removed assertions in ComposerCommandMenu.test.tsx, ComposerPrimaryActions.test.tsx, and MessagesTimeline.test.tsx fall squarely in the prohibited category, and none of the corresponding runtime components (ComposerCommandMenu.tsx, ComposerPrimaryActions.tsx, MessagesTimeline.tsx) or their CSS are modified by this PR. No new comments were posted.
Details
Note
Your check run agent prompt is: .macroscope/check-run-agents/ui-consistency.md
More information about how Check Run Agents work can be found in our Help Center.
Scope reviewed: apps/web/src/**/*.ts, apps/web/src/**/*.tsx, apps/web/src/**/*.css.
Changed in-scope files are all removals of unreferenced exports (linkEnvironment.ts, DiffPanelShell.tsx, ProviderUpdateLaunchNotification.logic.ts, providerIconUtils.ts, desktopUpdate.logic.ts, composer-logic.ts, composerDraftStore.ts, keybindings.ts, lib/contextWindow.ts, lib/utils.ts, portDiscoveryState.ts, providerInstances.ts, providerModels.ts, themePalette.ts) plus test-only deletions. No shared-primitive, Tailwind ownership, theme-variant, scroll/virtualization, or environment-routing behavior is altered by the diff.
No stateful actions were taken this run; prior inline comments remain resolved.
🔍 Agent activity — 2 tool calls
💭 The head commit is still ea2122a, unchanged from before, so my prior analysis holds — no new code changes to review. All four threads have been resolved with author responses, including one citing AGENTS.md policy against static-markup assertions, so I shouldn't duplicate my earlier comments here. I need to check AGENTS.md line 107 to weigh that justification properly, and the final output still needs to conform to the required JSON schema despite the "All clear" default guidance.
1. git_tools — markup",
2. browse_code — AGENTS.md
💭 The repo's committed policy explicitly bans the class of tests I flagged—static markup assertions and mock callback wiring—so my prior findings align with that policy, and the author's dismissal with reasons is justified. Since the code hasn't changed since the prior run, there's nothing new to flag here. I'll conclude with a success result, no findings, and a short summary noting UI consistency is clear.
Agent Credits: 42 credits
Token usage: 4 uncached input · 49,957 cache-read input · 51,877 cache-write input · 1,715 output
Agent Credits may also include non-token charges from external tools such as web research.