Add missing 'variant' property to task tool metadata in prompt-effect test, required after upstream added model variant support to the task tool's ExecuteResult type.
* fix: allow ChatGPT login over OpenAI provider config
Let users with inherited or globally configured OpenAI API keys sign in with ChatGPT for Codex without first removing their existing provider setup.
* fix: satisfy provider login ci checks
Annotate Kilo-specific shared CLI changes and apply webview formatting so the existing ChatGPT login flow passes CI.
Strengthen the Kilo queue integration test to capture each LLM
request body and assert the second request ends with the queued
user message, not a prior-turn assistant tail. Update the skipped
upstream prompt-effect active-run test comment to reference #9492
and point at the Kilo-specific authoritative coverage.
The gateway wraps @openrouter/ai-sdk-provider internally, so the same
openrouter SDK-side reasoning handling applies. Without this, users
routing through api.kilo.ai still see DeepSeek 400 'reasoning_content
must be passed back' errors.
The task tool propagates each child session's total cost into the parent's tool-wrapper assistant message. Counting both root and child sessions in `kilo stats` would double-count that contribution. Filter `getAllSessions()` to root sessions (parent_id IS NULL), matching how the TUI session list and web sidebar already treat child sessions.
Add a check for UTF-16 byte order marks before the NUL/control-char
heuristic so that legitimate UTF-16 encoded files are not incorrectly
classified as binary. Also update Opus 4.7 transform test to expect
the `display: "summarized"` property in xhigh reasoning config.
Replace object spread (`{...props}`) with `mergeProps` when forwarding
props through the plugin Slot wrapper. Spreading props in SolidJS
evaluates each prop once at mount time, freezing reactive values like
`ref`, `visible`, `disabled`, and `on_submit`. This caused handlers to
bind against stale closures after overlay transitions, breaking Enter
submission on the session_prompt slot.
Add regression tests verifying the wrapper uses `mergeProps` (not
spread) and that reactive prop tracking is preserved at runtime.
When a session spawned subagents via the task tool, the displayed cost reflected only the parent's own LLM steps. Propagate each subagent's total up to the invoking assistant message so every UI sums the full tree.
Add unit tests for the Encoding namespace and extend the tool integration
suite with ApplyPatch delete and EditTool replaceAll paths for non-UTF-8
files.
Resolve conflicts in question/index.ts, session/prompt.ts, and KiloProvider.ts to pull in upstream's flattened namespace refactor while preserving the queued-messages feature: Question.dismissAll, KiloQuestion.guardFollowup on ask, and the mid-turn 'break out when a follow-up is queued' check in runLoop. VS Code abortSession keeps its queue-preserving signature (no queuedMessageIDs delete); new autocomplete settings helpers from main are wired in unchanged.
Set and restore `Flag.KILO_EXPERIMENTAL_LSP_TOOL` around the two
`.live` LSP spawn tests so they actually reach the `spawn()` call
instead of being short-circuited by the TsClient guard.
Add `filediff` field to the edit tool's permission-ask metadata so the
VS Code extension's PermissionDock can render inline diff previews.
Sort and deduplicate dependency entries in package.json. Include a
comprehensive test suite verifying filediff presence across new file
creation, existing file edits, and multi-line replacements.
- Fix import paths in kilocode server instance to use routes directory
- Convert relative imports to path aliases in experimental and session routes
- Fix variable name bug in fromModelsDevModel (m -> base) for patchKiloModel
- Add null guard for VITE_KILO_CHANNEL in titlebar before calling includes
- Simplify model schema by replacing safeEnum with direct Schema.Literals
- Remove unused fullSyncedSessions set from sync context
- Add blocking option to question ask interface
- Remove FileTime dependency from read-directory and session-compaction tests
- Remove unnecessary @ts-expect-error comments for Flag overrides in tests
- Fix Account import path in config-gitignore test
- Regenerate SDK with new session list params (projectID, worktrees),
deduplicated Console class, and updated timeout default to 300000ms
Strip a leading U+FEFF from text before prepending a UTF-8/UTF-16 BOM so
round-tripping content that already carries a BOM character emits exactly
one BOM. Also clarifies the changeset that new files default to UTF-8.
Main moved the Patch namespace out of packages/opencode/src/patch/index.ts
into patch/patch.ts and switched tool imports to the new
@opencode-ai/shared/filesystem and ../../src/tool barrels. Reapply the
encoding-aware read/write changes on top of the refactored patch.ts, and
update the integration test imports to match the new module layout.
Replace granular submodule imports with their parent barrel index
across snapshot diff source and related test files. This covers
util/log, util/filesystem, config/config, provider/provider,
tool/registry, tool/truncate, and the filesystem shared package.
Replace the state.cancel-based mid-turn injection with a break-after-stream
queue. When a new prompt arrives while the assistant is streaming, the current
LLM step now finishes cleanly, any pending suggest/question tool auto-dismisses
via Suggestion.dismissAll and the new Question.dismissAll, and runLoop breaks
out before the next LLM step via KiloSessionPromptQueue.hasFollowup. Queued
prompts still run in order, each getting a full turn unless a newer one
arrives during it.
Alex Alecu