When every compaction round still overflowed the model context, SessionPrompt.runLoop would keep calling compaction forever and report the turn as completed. Cap attempts at three per turn and surface exhaustion as a ContextOverflowError on the assistant message with TurnClose reason=error.
Mark new plan follow-up sessions busy as soon as the tab opens so the UI
shows work is still in progress while the handover summary is generated.
Clear the temporary busy state if that pre-loop handover phase is aborted
or fails before the normal prompt loop takes over.
Show the plan text right away in the new session tab without waiting
for the handover LLM call. The message part is created with plan+todos
as soon as the session opens; the handover section is upserted onto the
same part once the slow LLM call resolves.
Create the follow-up session before running the handover LLM call so
the extension's pendingFollowup SSE gate fires inside its 30s TTL. The
handover and todos are now injected into the already-live session
after it resolves, instead of blocking session creation.
patchModelsDevModel was not including the options field in its return
value, so the Object.assign in fromModelsDevModel left options as {} and
discarded the description fetched from the Kilo Gateway API.
When the plan agent pauses for the follow-up question, CLI users had no
way to send a free-text reply: the main prompt input is hidden while a
blocking question is active, and the "Type your own answer" row was
being forced off to avoid duplicating the VS Code prompt input. CLI now
shows the custom-answer row again, while VS Code keeps it hidden.
The prompt queue's scope() hid the user message injected by
PlanFollowup.inject() because its ID was newer than the queue target,
so the loop saw the same plan_exit messages and re-asked the question
in an infinite cycle. Retarget the queue after inject so the new
message is visible on the next iteration.
Restore custom: false on the plan follow-up question — the "Type your own answer" row was redundant because the main prompt already routes typed text as a question reply. Auto-submit single-question single-select option picks in the VS Code QuestionDock so the button behaves like the TUI instead of silently waiting for a second Submit click.
Add optional `labelKey`/`descriptionKey` to Question.Option and `questionKey`/`headerKey` to Question.Info, annotated with `kilocode_change` markers. Populate these keys from the plan follow-up so the "Ready to implement?" question, "Start new session", and "Continue here" buttons render in the sidebar language while the canonical English labels remain on the reply wire (unchanged server-side matching).
* fix: persist custom provider model and variant deletions
The CLI config.update endpoint deep-merges its payload with existing
global config, so removing a model or reasoning variant from a custom
provider in the UI had no effect on disk. Save payloads now emit null
sentinels for removed IDs, and the Provider schema accepts nullable
record values so stripNulls can delete them during the merge.
Closes#9186
* chore: update kilo-vscode visual regression baselines
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Restore custom: false on the plan follow-up question — the "Type your own answer" row was redundant because the main prompt already routes typed text as a question reply. Auto-submit single-question single-select option picks in the VS Code QuestionDock so the button behaves like the TUI instead of silently waiting for a second Submit click.
Lock in that a session with an open suggestion reports as idle (so reopening VS Code or switching worktrees does not show it stuck/running), that accepting flips it back to busy without an idle flash, and that a dismissed suggestion leaves it idle for the run loop to resume cleanly.
Move worktree-scoped gitdir construction, ACP guard logic, and diff
caching directly into the core snapshot service, eliminating the
separate kilocode/snapshot module. This reduces indirection and keeps
all snapshot concerns co-located in a single file.
If deriveAndSyncStatus hangs, the per-session promise chain would never
drain and map entries would accumulate. Wrap the derive with a 3s timeout
so stuck work fails fast, is logged, and the existing cleanup runs.
The suggest tool blocks on a promise that only resolves when the user accepts or dismisses a suggestion. While blocked, the runner stayed in Running state and the session status remained busy — if the suggestion was never answered (e.g. VS Code was closed), the session was stuck forever and follow-up prompts appeared queued.
Mark the session idle while awaiting the user's response; the loop will set it back to busy when the suggestion resolves and processing continues.
Fixes#9150
Alex Alecu