feat(desktop): TODO/Schedule で Claude model と effort を指定可能に (closes #235)#245
Merged
feat(desktop): TODO/Schedule で Claude model と effort を指定可能に (closes #235)#245
Conversation
) TODO セッションとスケジュールの作成時に `--model` / `--effort` を 選べるようにした。これまではモデル・思考量の固定値で動いており、 VSCode の公式 Claude Code 拡張のような調整手段がなかった。 - TodoModal / ScheduleEditorDialog / 設定タブに共通ピッカーを追加 - supervisor が per-session の model / effort を CLI 引数に展開 - 設定タブに既定値を用意し、新規作成フォームにシードされる - DB: todo_sessions / todo_schedules に claude_model / claude_effort を追加
|
Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits. |
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 39 minutes and 49 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (5)
📝 WalkthroughWalkthroughこのプルリクエストは、TodoエージェントシステムにClaudeモデルと試行努力レベルの設定機能を追加します。データベーススキーマを拡張し、バックエンドの永続化層・オーケストレーション層・TRPC APIを更新し、新しいUIコンポーネントをフロントエンドに統合することで、ユーザーがスケジュール・セッション・デフォルト設定ごとにClaudeの実行パラメータを指定できるようにします。 Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx (1)
119-136:⚠️ Potential issue | 🟡 Minor新規スケジュール作成時にグローバル既定値がシードされていません。
TodoModal 側では
todoSettings.defaultClaudeModel/defaultClaudeEffortを初期値として反映していますが、この Dialog の新規作成ブランチではDEFAULT_SENTINELに固定されています。PR 概要の「new forms are seeded from those defaults」に反しており、スケジュール作成画面を開くと既定値が効かない UX になります。electronTrpc.todoAgent.settings.getを取得して TodoModal と同様にシードする実装が一貫します。🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx` around lines 119 - 136, The new-schedule branch in ScheduleEditorDialog is seeding Claude model/effort with DEFAULT_SENTINEL instead of the user's global defaults; update the "else" new-item initialization to read the todo settings (via electronTrpc.todoAgent.settings.get or the same prop/state TodoModal uses) and call setClaudeModel(todoSettings.defaultClaudeModel) and setClaudeEffort(todoSettings.defaultClaudeEffort) (falling back to DEFAULT_SENTINEL only if those settings are undefined), ensuring the new form fields are seeded consistently with the global defaults.
🧹 Nitpick comments (4)
apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/ClaudeRuntimePicker.tsx (1)
62-62:cnの単一引数呼び出しは冗長です。
cn("flex flex-col gap-1.5")は条件分岐もクラス合成もないため、直接className="flex flex-col gap-1.5"で十分です。細かいですが、cn利用箇所の意図が分かりやすくなります。♻️ Proposed fix
- <div className={cn("flex flex-col gap-1.5")}> + <div className="flex flex-col gap-1.5">🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/ClaudeRuntimePicker.tsx` at line 62, Replace the redundant single-argument cn call on the div in ClaudeRuntimePicker (the JSX element with className={cn("flex flex-col gap-1.5")}) with a plain string literal className; i.e., change className={cn("flex flex-col gap-1.5")} to className="flex flex-col gap-1.5" so the intent is clearer and unnecessary helper usage is removed.apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx (1)
25-34: tsconfig のエイリアス利用を検討してください。
../../../../ClaudeRuntimePickerの 4 階層を上る相対パスは、ファイルを別階層へ移動したときに壊れやすいです。renderer/features/todo-agent/ClaudeRuntimePickerなど、tsconfig で定義済みのエイリアスに寄せると保守しやすくなります。As per coding guidelines, "Use aliases as defined intsconfig.jsonwhen possible".🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx` around lines 25 - 34, The import uses a fragile deep relative path; update the import of ClaudeRuntimePicker and its symbols (ClaudeRuntimePicker, DEFAULT_SENTINEL, ClaudeModelPick, ClaudeEffortPick, fromPersistedEffort, fromPersistedModel, toPersistedEffort, toPersistedModel) to use the tsconfig path alias instead of "../../../../ClaudeRuntimePicker" (e.g., the alias that maps to the renderer/features/todo-agent/ClaudeRuntimePicker), and ensure your tsconfig/paths is configured and tests/build still resolve the new alias.apps/desktop/src/main/todo-agent/types.ts (1)
28-39: 固定モデル名のハードコードは陳腐化しやすい点に注意。
claude-opus-4-7/claude-sonnet-4-6/claude-haiku-4-5-20251001は CLI 側の実モデル名に依存しており、新リリースのたびにこの enum と UI ラベル（claudeRuntimeOptions.ts）の両方を更新する必要があります。PR 記述にもある通り model×effort の互換性検証はアプリ側で行わないため、古い固定名を選んだユーザーは起動時に Claude CLI 側でセッション失敗します。最低限、固定名に関しては「使えなくなる可能性がある」旨の注意を UI の description か README に残し、エイリアス（
opus/sonnet/haiku）の使用を推奨しておくと運用が楽です。コード上の対応は不要ですが、supervisor 側で spawn 失敗時のエラーメッセージに「モデル名が無効な可能性があります」のヒントを含めることを検討してください。🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/desktop/src/main/todo-agent/types.ts` around lines 28 - 39, The hardcoded Claude model names (CLAUDE_MODEL_OPTIONS / TodoClaudeModel / todoClaudeModelSchema) can become invalid across CLI releases; update the UI description and repository README to warn users that concrete model strings (e.g., "claude-opus-4-7") may break and recommend using the stable aliases ("opus", "sonnet", "haiku") instead, and also add a hint in the supervisor/error path (where spawn/session failures are logged) to mention that an invalid model name could be the cause so users know to try the alias or update the model string.apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/claudeRuntimeOptions.ts (1)
19-22:DEFAULT_SENTINELが将来のモデル/effort 値と衝突しないことを保証する型レベルのガードを検討。
"__default__"は現状のCLAUDE_MODEL_OPTIONS/CLAUDE_EFFORT_OPTIONSと衝突しませんが、将来誰かが options に"__default__"を追加すると、ClaudeModelPickのユニオンが縮退して静かにバグります。以下のような静的アサートを 1 行足しておくと安全です。♻️ 提案
export const DEFAULT_SENTINEL = "__default__" as const; +// Compile-time guard: sentinel must not collide with any real option. +type _AssertSentinelUnique = + typeof DEFAULT_SENTINEL extends TodoClaudeModel | TodoClaudeEffort + ? never + : true;🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/claudeRuntimeOptions.ts` around lines 19 - 22, DEFAULT_SENTINEL が将来の options 値と衝突してユニオンが縮退しないよう、型レベルのガードを追加してください: 具体的には DEFAULT_SENTINEL、ClaudeModelPick、ClaudeEffortPick を用いるファイルで CLAUDE_MODEL_OPTIONS と CLAUDE_EFFORT_OPTIONS に対してコンパイル時の否定的アサーション（conditional type / never を用いた静的チェック）を追加し、DEFAULT_SENTINEL がいずれの options 値にも割り当てられないことを保証してください（エラーが出るようにして将来の衝突を防ぐ）。
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/claudeRuntimeOptions.ts`:
- Around line 134-150: fromPersistedModel and fromPersistedEffort silently map
unknown persisted strings to DEFAULT_SENTINEL; update both functions to emit a
warning when they encounter a persisted value not present in
CLAUDE_MODEL_OPTIONS / CLAUDE_EFFORT_OPTIONS (e.g., call console.warn or the
shared logger) including the offending persisted value and which function
detected it, and keep returning DEFAULT_SENTINEL; optionally add an overload or
callback parameter to surface this event to callers (TodoModal /
ScheduleEditorDialog / PresetsDialog) if they need to notify users.
In
`@apps/desktop/src/renderer/features/todo-agent/TodoManager/PresetsDialog/PresetsDialog.tsx`:
- Around line 119-130: The current useEffect unconditionally overwrites all
local fields whenever settings changes, which will clobber user edits during
background refetch; update the effect in PresetsDialog to only sync settings
when the form is not dirty or only on initial mount: add a dirty flag (or an
initialized ref) and wrap the
setMaxIter/setMaxMin/setMaxConcurrent/setRetentionDays/setDefaultModel/setDefaultEffort
calls inside an if (!dirty) / if (!initialized) check, or change the effect to
run only once on mount and provide separate explicit sync logic for saves/loads
so in-progress edits are preserved (refer to useEffect, settings, and the setter
functions named above).
In `@apps/desktop/src/renderer/features/todo-agent/TodoModal/TodoModal.tsx`:
- Around line 87-103: The effect re-seeds state whenever todoSettings updates,
which can overwrite a user-selected DEFAULT_SENTINEL; modify the useEffect
containing setClaudeModel and setClaudeEffort so it only runs once on initial
mount or only on the first arrival of todoSettings (e.g., track a seeded flag
via useRef or detect that todoSettings was previously undefined) instead of on
every todoSettings change; keep the existing DEFAULT_SENTINEL checks but gate
the whole seeding logic with the seeded flag to prevent overwriting a deliberate
DEFAULT_SENTINEL selection by the user.
In `@packages/local-db/src/schema/todo-sessions.ts`:
- Around line 89-99: The supervisor's runClaudeTurn is currently pushing
params.claudeModel and params.claudeEffort into the spawn args without whitelist
checks; add validation there using the same logic as
claudeRuntimeOptions.fromPersistedModel and .fromPersistedEffort (or directly
check against CLAUDE_MODEL_OPTIONS and CLAUDE_EFFORT_OPTIONS) and only push
"--model" / "--effort" when the value is present and matches the allowed option
set; do not pass through arbitrary DB values—reject or skip unknown values so
spawn() only receives validated model/effort arguments.
---
Outside diff comments:
In
`@apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx`:
- Around line 119-136: The new-schedule branch in ScheduleEditorDialog is
seeding Claude model/effort with DEFAULT_SENTINEL instead of the user's global
defaults; update the "else" new-item initialization to read the todo settings
(via electronTrpc.todoAgent.settings.get or the same prop/state TodoModal uses)
and call setClaudeModel(todoSettings.defaultClaudeModel) and
setClaudeEffort(todoSettings.defaultClaudeEffort) (falling back to
DEFAULT_SENTINEL only if those settings are undefined), ensuring the new form
fields are seeded consistently with the global defaults.
---
Nitpick comments:
In `@apps/desktop/src/main/todo-agent/types.ts`:
- Around line 28-39: The hardcoded Claude model names (CLAUDE_MODEL_OPTIONS /
TodoClaudeModel / todoClaudeModelSchema) can become invalid across CLI releases;
update the UI description and repository README to warn users that concrete
model strings (e.g., "claude-opus-4-7") may break and recommend using the stable
aliases ("opus", "sonnet", "haiku") instead, and also add a hint in the
supervisor/error path (where spawn/session failures are logged) to mention that
an invalid model name could be the cause so users know to try the alias or
update the model string.
In
`@apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/claudeRuntimeOptions.ts`:
- Around line 19-22: DEFAULT_SENTINEL が将来の options
値と衝突してユニオンが縮退しないよう、型レベルのガードを追加してください: 具体的には
DEFAULT_SENTINEL、ClaudeModelPick、ClaudeEffortPick を用いるファイルで CLAUDE_MODEL_OPTIONS
と CLAUDE_EFFORT_OPTIONS に対してコンパイル時の否定的アサーション（conditional type / never
を用いた静的チェック）を追加し、DEFAULT_SENTINEL がいずれの options
値にも割り当てられないことを保証してください（エラーが出るようにして将来の衝突を防ぐ）。
In
`@apps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/ClaudeRuntimePicker.tsx`:
- Line 62: Replace the redundant single-argument cn call on the div in
ClaudeRuntimePicker (the JSX element with className={cn("flex flex-col
gap-1.5")}) with a plain string literal className; i.e., change
className={cn("flex flex-col gap-1.5")} to className="flex flex-col gap-1.5" so
the intent is clearer and unnecessary helper usage is removed.
In
`@apps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsx`:
- Around line 25-34: The import uses a fragile deep relative path; update the
import of ClaudeRuntimePicker and its symbols (ClaudeRuntimePicker,
DEFAULT_SENTINEL, ClaudeModelPick, ClaudeEffortPick, fromPersistedEffort,
fromPersistedModel, toPersistedEffort, toPersistedModel) to use the tsconfig
path alias instead of "../../../../ClaudeRuntimePicker" (e.g., the alias that
maps to the renderer/features/todo-agent/ClaudeRuntimePicker), and ensure your
tsconfig/paths is configured and tests/build still resolve the new alias.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: ec3c8319-e99d-4d55-ac0f-09101ccfa7f0
📒 Files selected for processing (18)
apps/desktop/src/main/todo-agent/schedule-store.tsapps/desktop/src/main/todo-agent/scheduler.tsapps/desktop/src/main/todo-agent/session-store.tsapps/desktop/src/main/todo-agent/settings.tsapps/desktop/src/main/todo-agent/supervisor.tsapps/desktop/src/main/todo-agent/trpc-router.tsapps/desktop/src/main/todo-agent/types.tsapps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/ClaudeRuntimePicker.tsxapps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/claudeRuntimeOptions.tsapps/desktop/src/renderer/features/todo-agent/ClaudeRuntimePicker/index.tsapps/desktop/src/renderer/features/todo-agent/TodoManager/PresetsDialog/PresetsDialog.tsxapps/desktop/src/renderer/features/todo-agent/TodoManager/SchedulesSection/components/ScheduleEditorDialog/ScheduleEditorDialog.tsxapps/desktop/src/renderer/features/todo-agent/TodoModal/TodoModal.tsxpackages/local-db/drizzle/0058_todo_claude_model_effort.sqlpackages/local-db/drizzle/meta/0058_snapshot.jsonpackages/local-db/drizzle/meta/_journal.jsonpackages/local-db/src/schema/todo-schedules.tspackages/local-db/src/schema/todo-sessions.ts
Owner
Author
|
@codex review |
- supervisor が --model / --effort をスプロウンに渡す前に CLAUDE_*_OPTIONS 相当のホワイトリスト検証を行う。DB が破損して 不正値が入っても spawn 引数に漏れないよう二重の防御層にする - fromPersistedModel / fromPersistedEffort が未知値を黙って DEFAULT_SENTINEL に落とす挙動は維持しつつ、DevTools に console.warn を残してサイレント退行を観測できるようにする - TodoModal と PresetsDialog の useEffect を useRef ガードで 初回のみ実行にする。React Query のバックグラウンド refetch や ユーザが明示的に選んだ「デフォルト」の上書きを防ぐ
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 13f6428b67
ℹ️ About Codex in GitHub
Your team has set up Codex to 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 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Codex P2 指摘。これまで ScheduleEditorDialog の新規作成モードでは model/effort が常に DEFAULT_SENTINEL に初期化されており、TodoModal と 違って設定タブの既定値が反映されなかった。結果として新規スケジュール 保存時にも null が書かれ、発火時の session にも null がコピーされる ため --model/--effort が渡らず、既定値が事実上効かない状態だった。 - todoAgent.settings.get を ScheduleEditorDialog にも購読 - initial === null のときは claudeSeededRef 越しに一度だけ既定値で モデル/effort をシード。React Query refetch での再シードを防ぎ かつユーザの「デフォルト」選択を保護する
MocA-Love
added a commit
that referenced
this pull request
Apr 17, 2026
…chment を daemon へ移植 main 取り込み側で supervisor.ts に加わった以下の機能を、daemon 方式の 新しいアーキテクチャ (supervisor-engine.ts) に再実装する。 #244 ScheduleWakeup 継続 - engine に `wakeupResumeMarkers` を追加し、scheduler からの `fromScheduledWakeup: true` を経由してマーカーを立てる - `start()` option をIPC プロトコル (StartRequest) に追加し、 supervisor.ts (proxy) → client.ts → daemon/index.ts → engine.start の順にフラグを伝搬 - `runSession` がマーカーを消費し「再開」バナー/previous text 再送を スキップして短い継続キューだけを送る - `buildIterationPrompt` に `isScheduledWakeupContinuation` を追加、 claudeSessionId が null のときは iteration-1 フルプロンプトへ フォールバック #245 Claude model/effort - `runClaudeTurn` に `claudeModel`/`claudeEffort` を渡し、 CLAUDE_*_OPTIONS ホワイトリストで検証してから `--model`/`--effort` を spawn に追加（不明値は warn して無視する二重防御） - setup event に「Claude 設定」として選択内容を表示 #250 画像添付 tool_result - `extractToolResultDetails` が image ブロックのみの tool_result も `[画像 N 件]` プレースホルダで emit するよう修正 - `extractAttachmentPaths` を daemon 側にも複製し、iteration 1 の プロンプトに添付ファイルパスを列挙して Read を促す #243 done→resume は trpc-router.ts 側で完結するので追加変更なし
This was referenced Apr 22, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
claude -pにそのまま--model/--effortを渡す形式なので、公式 CLI がサポートする alias (opus/sonnet/haiku) と固定名 (claude-opus-4-7/claude-sonnet-4-6/claude-haiku-4-5-20251001)、effort はlow/medium/high/xhigh/maxを選べるCloses
Changes
Model / Effort に関する注意
Test plan
Summary by CodeRabbit
リリースノート