Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions devlog/_plan/260908_bug6_manual_stack/030_v2_guidance.md
Original file line number Diff line number Diff line change
Expand Up @@ -47,3 +47,7 @@ Remote-only focused activation: `bun test tests/codex-integration/multi-agent-co


Main decision: preserve the complete original diff. The optional extra server caller fixture is deferred unless source audit reveals an untested change; do not duplicate the existing replay matrix merely for volume. Sync structure/03_catalog-and-subagents.md to the new tag and policy boundary.

## wp3 P refresh

Previous wp2 D: PR3991 head00eb47886 passed run34180674115, source audit and remote docs425pages; proceed guidance carry. Prepared layer3 consists of24977adf2,21757b71a,8000e2482, based on d1f61e933. Intervening wp2 changes affect xAI adapter/tests, provider/adapters docs and structure04; none overlap the9layer3 files. Original #3944 remains open at6fb0fc6f. Independent prepared-source/security audit PASS in isolated v2GuidanceReviewer.md; actual adoption requires unchanged-delta/interdiff verification and own hostedCI.
18 changes: 18 additions & 0 deletions devlog/_plan/260908_bug6_manual_stack/040_v2_preset.md
Original file line number Diff line number Diff line change
Expand Up @@ -74,3 +74,21 @@ Suggested main-owned scratch outputs: `ui-v2/source-manifest.json`, `ui-v2/reque


Main decision: preserve original 23-file diff and all regressions. Suggested extra CLI/API test cases remain a targeted audit decision rather than mandatory scope expansion. Sync structure/03_catalog-and-subagents.md and structure/05_gui-and-management-api.md with the response-only recommendation and explicit-write behavior. Render QA uses the existing built artifact.

## wp4 P refresh and artifact publication

Previous wp3 D: PR3992 at3ceef0121 passed CI34181398746, target-check rerun34181398713, independent source/security audit and425page remote docs build. Proceed server-owned preset. Candidate5b60a4fa7+e5a6f35b8 was independently audited25files; all25preimage files on actual predecessor3ceef0121 equal candidatebase8000e2482. Preserve original contribution and both layers of agent guides.

UI proof uses the immutable matching hosted dashboard artifact and reviewed auxiliary stdlib fixture at /tmp/ocx-bug6-prep-v2-01a07e9d/.tmp/ui-v2/server.py. No product server/build or actual account operations. Capture and inspect representative desktop/mobileEN/KO, valid/custom/restore/save/clear/missing/malformed/error cases; hosted tests retain in-place delayed-server-switch coverage. Publish only sanitized screenshot assets on an owned evidence branch via noverify push and link immutable commit URLs in this PR, so adding review evidence does not rewrite the tested product head. This evidence-only branch is not another product PR or native stack. Actual product delivery remains the one six-layer chain. Preserve asset provenance/buildGUI-tree checks and teardown proof.

## C render foldback: narrow editor layout

Real prebuilt GUI capture at320px shows an unbroken configuration-key hint clipped and the textarea compressed beside both actions;390px has no page overflow but editing space is unnecessarily narrow. Evidence ui-qa/ko-320-custom.png and ko-mobile-custom.png under ignored scratch. This is a bounded improvement of the changed preset surface, not a claim that the original PR introduced the old layout.

MODIFY gui/src/styles-subagents-workspace.css only: add overflow-wrap:anywhere to existing delegation setting hints; give the preset editor a minmax(0,1fr)/auto/auto grid with min-width0 textarea; reuse existing640px media breakpoint to span textarea across a full first row and place two actions on the second row. NoJS, labels, fields or authority semantics change. Rebuild only remotely; compare head-stamped320/390/1440captures and DOMwidth/textareawidth, plus existing frontendgates. Retain prior functional UI request evidence becauseJS unchanged; verify a representative Save on finalrender. Main re-audits this CSS scope before B and captures finalartifact before publishing screenshotproof.

## C verification foldback: bounded guard probes

Run34183701289 timed out on macos1/2 after test-home-guard passed its unregistered-home case, then Bun reported a dangling process. Runtime root cause remains unproven (H1retainedchildhandles/H2syncwait/H3isolate state); source establishes an unbounded synchronous probe owner. This is a verification-harness correction, not a claim to fix Bun internals. No production guard behavior changes.

MODIFY tests/ci-workflows/test-home-guard.test.ts only: convert runProbe and all its callers to awaited async Bun.spawn; reuse exported captureTestOutput from scripts/test.ts (its executable entry is import.meta.main gated, already imported by test-runner tests); preserve exact argv/cwd/environment filtering/fake homes and all existing assertions. Use existing watchdogMs(5000) below CI60s test ceiling, bounded execution/TERM/KILL/reap/output-drain stages with cleared timers and cleanup of probe scratch. Require natural exit0, no signal, complete output; timeout/nonzero/incomplete/unreaped is a thrown diagnostic, never a success marker. Record safe probe-boundary ID/PID/exit diagnostics without inherited environment. Add synthetic nonzero-exit and nonterminating-probe controls proving failures are reported and owned child gone; do not skip guards or extend CI deadlines. All product verification stays remote. If new diagnostics still stall/fail, investigate that boundary; passing the converted harness is not an upstream-root-cause claim.
14 changes: 12 additions & 2 deletions docs-site/src/content/docs/ja/reference/configuration/agents.md
Original file line number Diff line number Diff line change
Expand Up @@ -26,15 +26,25 @@ description: マルチエージェント サーフェス、委任ガイダンス

管理 API は、`GET`/`PUT /api/v2`、`/api/injection-model`、`/api/effort-caps`、`/api/subagent-models`、および `/api/subagent-model-fallback` を公開します。インジェクションモデルの更新は部分的です。カスタム プロンプトは、その API の `prompt` フィールドです。

## 常時プロアクティブ委任

Subagents → 詳細設定の **常時プロアクティブ委任**(旧 **Ultra mode**)は、推論のエフォートを変更せず、委任を開始する条件だけを変更します。推奨プリセットでも、ユーザーの指示、権限の境界、タスクの範囲、ツールの規則は維持されます。

`GET` と `PUT /api/v2` は、追加で `multiAgentModeHintRecommendation: { text, revision }` を返します。ダッシュボードはプリセットの有効化と復元にサーバー提供のテキストを使い、固定文へのフォールバックは行いません。旧サーバーが推奨値を返さない場合や値の形式が不正な場合、プリセットの適用と復元は利用できませんが、既存のカスタムヒントの編集と削除は引き続き利用できます。プリセットの復元はローカルの下書きだけを変更し、保存操作で確定します。

設定の読み取り、無関係な更新、アップグレードでは、保存済みのヒントは移行されません。ヒントを明示的に更新し、その本文が既知の旧 OpenCodex プリセット2種類のいずれかとバイト単位で完全一致する場合に限り、現在の推奨文に置き換わります。それ以外の有効なカスタムテキストは、空白だけが異なるものも含め、バイト単位で保持されます。既存の v2 有効化、機能サポート確認、ヒント削除の規則は変わらず、変更は新しい Codex セッションに適用されます。

## ロスターとガイダンス

有効な v2 ロスターは、v2 と互換性があり、挿入されたカタログに存在する、構成され、ピッカーに表示され、優先順位で並べ替えられた最初の 5 つのモデルです。 V2 の適格性は、明示的な `"v2"`、`null`、または欠落しているアップストリーム ピンを適格なものとして扱います。実際の `"v1"` ピンは除外されます。除外されたエントリは設定に残るため、後で適格になる可能性があります。
有効な v2 ロスターは、設定済みでピッカーに表示され、優先順位で並べ替えられた最初の 5 つのモデルのうち、挿入されたカタログに存在し、明示的に `"disabled"` とされていないモデルです。明示的な `"v2"` ピンは再帰的なワーカーをサポートし、`"v1"`、`null`、ピンの省略はリーフワーカーとして引き続き適格です。除外されたエントリは設定に残るため、後で適格になる可能性があります。

表面検出はツール形状を使用します。 `send_input`、`resume_agent`、または `close_agent` を持つ名前空間付き `spawn_agent` は v1 です。 `send_message`、`followup_task`、`interrupt_agent`、または `list_agents` を備えたフラット `spawn_agent` は v2 です。

V1 ガイダンスは、`max` または `ultra` でのみプロアクティブ テキストです。 V2 は、優先モデル、適格なロスター、またはフォールバック チェーンが存在する場合にのみ、プロキシ作成の開発者メッセージを受信します。組み込みの v2 ガイダンスには 700 文字のバジェットがあり、必要に応じて最初にロスターが削除されます。ガイダンスはリプレイ プレフィックス全体で重複排除され、後続の `compaction_trigger` の前に挿入されます。

`injectionModel` および `injectionEffort` は、ネイティブデフォルト同期が有効になっていない限り、推奨事項です。組み込みの v2 テキストは、サポートされているモデル/エフォートのオーバーライドを `fork_turns: "none"` を使用して `spawn_agent` に渡すように Codex に要求します。カスタム `injectionPrompt` は、欠落している値を空の文字列に置き換えます。
組み込みの v2 サブエージェントガイダンスとカスタム `injectionPrompt` 本文は、どちらも `<opencodex_subagent_guidance>` を使用し、Codex ネイティブの `<multi_agent_mode>` メッセージとは区別されます。組み込みテキストは、解決済みの優先モデル、ロスター、フォールバックチェーンを示しますが、委任、モデルのオーバーライド、`fork_turns` は指示しません。カスタム本文のプレースホルダー置換と内容は維持されます。`injectionModel` および `injectionEffort` は、ネイティブデフォルト同期が有効になっていない限り推奨事項であり、カスタムプレースホルダーの欠落値は引き続き空の文字列に置き換えられます。

リプレイの重複排除では、タグの種類ごとに最新のテキストとの完全一致を確認します。両方の値が新しいプロキシのタグを使用する場合、カスタムガイダンスから組み込み形式へ戻すと、その時点の内容が追加されます。途中でネイティブモードが変わっても、変更のないプロキシガイダンスは重複追加されません。既存のネイティブメッセージと旧タグ付きの履歴は保持されます。ラッパーの変更によって過去のメッセージの作成者が判明したり、以前の指示が取り消されたりするわけではありません。複数バージョンが混在する履歴は、旧タグだけでは分類できず、そのような履歴での設定変更の検出は保証されません。

## ネイティブ Codex のデフォルト同期

Expand Down
14 changes: 12 additions & 2 deletions docs-site/src/content/docs/ko/reference/configuration/agents.md
Original file line number Diff line number Diff line change
Expand Up @@ -26,15 +26,25 @@ description: 멀티 에이전트 표면, 위임 안내, 선호 모델, 대체

관리 API는 `GET`/`PUT /api/v2`, `/api/injection-model`, `/api/effort-caps`, `/api/subagent-models`, `/api/subagent-model-fallback`를 제공합니다. injection-model 업데이트는 부분 업데이트입니다. 사용자 지정 프롬프트는 이 API의 `prompt` 필드입니다.

## 항상 능동 위임

서브에이전트 → 고급의 **항상 능동 위임**(이전 이름: **울트라 모드**)은 추론 노력을 바꾸지 않고 위임을 시작하는 조건만 변경합니다. 권장 프리셋에서도 사용자 지침, 권한 경계, 작업 범위, 도구 규칙은 유지됩니다.

`GET`과 `PUT /api/v2`는 `multiAgentModeHintRecommendation: { text, revision }`도 반환합니다. 대시보드는 프리셋을 켜거나 복원할 때 서버가 제공한 텍스트를 사용하며, 고정된 대체 문구는 사용하지 않습니다. 이전 서버가 권장값을 제공하지 않거나 잘못된 형식으로 반환하면 프리셋 적용·복원은 사용할 수 없으며, 기존 사용자 지정 힌트의 편집·삭제는 계속 사용할 수 있습니다. **프리셋 복원**은 로컬 초안만 바꾸고, **저장**을 눌러야 저장됩니다.

설정 조회, 관련 없는 업데이트, 업그레이드는 저장된 힌트를 자동 변경하지 않습니다. 힌트를 명시적으로 업데이트할 때 본문이 알려진 두 가지 이전 OpenCodex 프리셋 중 하나와 바이트 단위로 정확히 일치하는 경우에만 현재 권장 문구로 바뀝니다. 그 밖의 유효한 사용자 지정 텍스트는 공백 차이가 있는 변형까지 바이트 단위로 보존됩니다. 기존 v2 활성화·기능 지원 확인·힌트 해제 규칙은 그대로이며, 변경은 새 Codex 세션에 적용됩니다.

## 로스터와 안내

실제 v2 로스터는 설정되어 있고, 선택기에 보이며, 우선순위로 정렬된 상위 다섯 모델 중 v2와 호환되고 주입된 카탈로그에 존재하는 모델입니다. v2 적격성은 명시적인 `"v2"`, `null`, 또는 생략된 상위 고정값을 적격으로 보고, 실제 `"v1"` 고정값은 제외합니다. 제외된 항목은 나중에 적격이 될 수 있도록 설정에 그대로 남습니다.
실제 v2 로스터는 설정되어 있고, 선택기에 보이며, 우선순위로 정렬된 상위 다섯 모델 중 주입된 카탈로그에 존재하고 명시적으로 `"disabled"`로 표시되지 않은 모델입니다. 명시적인 `"v2"` 고정값은 재귀 작업자를 지원하며, `"v1"`, `null`, 생략된 고정값도 하위 작업을 다시 위임하지 않는 작업자로 참여할 수 있습니다. 제외된 항목은 나중에 적격이 될 수 있도록 설정에 그대로 남습니다.

표면 판별은 도구 형태를 기준으로 합니다. 네임스페이스가 붙은 `spawn_agent`에 `send_input`, `resume_agent`, `close_agent`가 있으면 v1입니다. 평평한 `spawn_agent`에 `send_message`, `followup_task`, `interrupt_agent`, `list_agents`가 있으면 v2입니다.

V1 안내는 `max` 또는 `ultra`에서만 선제 텍스트로 제공됩니다. V2는 선호 모델, 적격 로스터, 대체 체인 중 하나가 있을 때만 프록시가 작성한 개발자 메시지를 받습니다. 내장 v2 안내에는 700자 예산이 있고, 필요하면 로스터를 먼저 줄입니다. 안내는 replay prefix 전반에서 중복 제거되며, 뒤에 오는 `compaction_trigger` 앞에 삽입됩니다.

`injectionModel`과 `injectionEffort`는 네이티브 기본값 동기화가 활성화되지 않으면 권고 수준입니다. 내장 v2 텍스트는 Codex에게 지원되는 모델/노력 오버라이드를 `fork_turns: "none"`과 함께 `spawn_agent`로 전달하라고 요청합니다. 사용자 지정 `injectionPrompt`는 누락된 값을 빈 문자열로 대체합니다.
내장 v2 서브에이전트 안내와 사용자 지정 `injectionPrompt` 본문은 모두 `<opencodex_subagent_guidance>`를 사용하며, Codex 네이티브 `<multi_agent_mode>` 메시지와 구분됩니다. 내장 텍스트는 결정된 선호 모델, 모델 목록, 대체 체인을 알리지만 위임, 모델 오버라이드, `fork_turns`를 지시하지는 않습니다. 사용자 지정 본문의 자리표시자 치환과 내용은 유지됩니다. `injectionModel`과 `injectionEffort`는 네이티브 기본값 동기화가 활성화되지 않으면 계속 권고 수준이며, 사용자 지정 자리표시자의 누락된 값은 빈 문자열로 대체됩니다.

replay 중복 제거는 각 태그 계열의 가장 최근 텍스트와 정확히 일치하는지 비교합니다. 두 값 모두 새 프록시 태그 계열을 사용하는 경우, 사용자 지정 안내에서 내장 형식으로 돌아오면 현재 안내가 추가됩니다. 그 사이에 네이티브 모드가 바뀌어도 변경되지 않은 프록시 안내가 중복 추가되지는 않습니다. 기존 네이티브 메시지와 예전 태그가 붙은 이력은 보존됩니다. 래퍼 변경으로 과거 메시지의 작성자가 판별되거나 이전 지침이 철회되는 것은 아닙니다. 여러 버전이 섞인 이력은 예전 태그만으로 분류할 수 없으며, 이러한 이력에서 설정 전환이 감지된다고 보장하지 않습니다.

## Codex 기본값 동기화

Expand Down
Loading
Loading