computeNonMessageBreakdown estimated Skills tokens from the unfiltered
session.skills registry and subtracted that from the first system-prompt
block, which only ever contained the rendered (filtered) skills. Hidden
explicit-only skills (hide/disable-model-invocation) and all skills when
the read tool was absent inflated Skills and clamped System prompt to 0.
Add a renderedSkills helper mirroring buildSystemPrompt's filter and use
it for the Skills estimate so the category split matches the provider
prompt.
Fixes#6498
- Add `pinSessionOAuthAccount` storage method and active account flag to the auth storage API.
- Introduce session account selector component, controller logic, and interactive mode delegation.
- Implement the `/session pin` builtin slash command with text listing and account pinning capabilities.
- Add unit tests covering session account selection, component navigation, and command handling.
Scheduling agent.continue() inside navigateTree started a post-prompt
task before the interactive /tree handler rebuilt its transcript, so a
fast provider's agent_start/turn_start could render against the stale
pre-rebuild UI and then be clobbered by renderInitialMessages.
navigateTree now reports the commit via askReanswerCommitted instead of
resuming, and SelectorController.showTreeSelector calls the new
AgentSession.resumeAfterAskReanswer() only after renderInitialMessages +
reloadTodos, so the resumed turn always renders against the rebuilt
transcript.
Addresses chatgpt-codex-connector review on #6484.
Fixes#6483
- Pinned displaceable snapshots like hub waiting polls and todo lists to the viewport during active execution.
- Prevented unpinned spinner ticks from repeatedly committing mutating rows to native scrollback and force-sealing blocks.
Moved vocalizer interruption from agent_start to user message_start so agentLoopContinue preserves queued speech.
Expanded regression coverage for continuation starts and user prompt boundaries.
Fixes#6375
- Replaced single-provider preferences with ordered priority lists for web search and image generation.
- Added a `MultiSelectSubmenu` component supporting toggle and reordering interactions in settings.
- Implemented migration logic to convert legacy single-provider preferences into ordered priority lists.
- Updated setup wizard scenes, image generation fallback logic, and search provider chains to use priority lists.
Moved vocalizer clearing from each assistant/tool continuation boundary to the start of a new agent run.
Added regression coverage for uninterrupted internal turns and new-run interruption.
Fixes#6375
Reverted branch-side edits to spawn-policy prompts/tests, settings tab
groups, mermaid cache typing, prewalk todo gating, and packages/ai test
churn back to merge-base content; trimmed their changelog entries. These
repaired stale CI against an older main and are stale or conflicting
against current main.
- Added transcriptMessageComponents to every InteractiveMode ctx test literal; #6033's reuse cache made the field required and addMessageToChat populates it unconditionally.
- Wired setTodoPhases into the eager-todo ToolSession fixture to mirror sdk.ts; the test relied on the stale message_end todo replay #6148 removed.
The fullscreen Plan Review overlay stayed mounted, focused, and fully
interactive after a choice was picked. showPlanReview's finish() resolves
the choice promise but deliberately defers the hide until #approvePlan
(guarded for #5688/#5319/#5689), so during slow async approval work — e.g.
context compaction — arrow keys still moved the cursor and repeat
Enter/Esc were silently swallowed by the settle guard, with the
fullscreen buffer masking all progress underneath. Users on Ghostty and
macOS Terminal read this as "/plan is frozen".
The overlay now flips to a committed state the moment a choice fires:
handleInput is a no-op, and the render shows a "<choice> — submitting…"
indicator plus an "applying your selection" footer. The deferred-hide
timing is untouched, so the existing stale-buffer and focus guards stay
intact.
Fixes#5926
Upstream's shift+enter pre-answer feature (merged via upstream/main) changed
TreeSelectorComponent's onSelect callback to accept a second (options)
parameter carrying summarize state. The pickEntry() test helper still called
onSelect with a single argument, so options.summarize threw undefined-object
errors once merged.
- tools.test.ts still asserted the pre-#5812 ±context expansion for offset/
limit and archive-entry reads; updated to the exact-bounds contract.
- selector-controller-logout.test.ts mocked the old modelRegistry.refresh;
#5786 switched logout to a provider-scoped refreshProvider(id, 'online'),
so the mock never resolved and the test timed out.
- Removed `#compactedChildStart` state, `compactable` segment field, and `#compactCommittedPrefix()` method, stopping the local-frame pruning of committed finalized rows.
- Post-finalize mutations now surface on every render via version tracking, instead of requiring a destructive replay to rehydrate the compacted block.
- Changed `print-mode.ts` to use `session.getLastAssistantMessage()` instead of array tail read, fixing access to terminal assistant messages after compaction removal.
- Kept tall-preview pages inside the selected option row.\n- Compared cached overflow output against the initial width-adjusted render.\n\nNote: pre-existing repository formatting failures in bun check are not addressed by this PR.
Ask dialogs accepted the configured cancel binding but always advertised Escape. Derive the compact footer label from the active binding so the displayed key remains actionable.
Rich ask options previously exposed preview content only for the highlighted row. Render every preview beneath its option and keep long content reachable without reviving the split-pane layout.
can1357