omnilane 0.7.0 → 0.7.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (9) hide show
  1. package/CHANGELOG.md +14 -1
  2. package/README.ja.md +12 -2
  3. package/README.ko.md +12 -2
  4. package/README.md +13 -2
  5. package/README.zh-CN.md +10 -2
  6. package/README.zh-TW.md +10 -2
  7. package/VERSION +1 -1
  8. package/package.json +1 -1
  9. package/routing.yaml +2 -2
package/CHANGELOG.md CHANGED
@@ -6,6 +6,18 @@ semantic version tags.
6
6
 
7
7
  ## [Unreleased]
8
8
 
9
+ ## [0.7.1] - 2026-07-18
10
+
11
+ ### Changed
12
+
13
+ - Routing defaults refreshed against 2026-07 model data (Artificial Analysis
14
+ Coding Agent Index v1.1, vendor announcements): `hardest-coding` first
15
+ choice moves from GPT-5.6 Sol `xhigh` to `max` — Sol (max) scores 80, the
16
+ current state of the art, retiring the earlier xhigh-beats-max snapshot.
17
+ - Claude Opus 4.8 fallbacks on `hardest-coding` and `hard-judgment` move from
18
+ `high` to `xhigh` effort, following Anthropic's published guidance to use
19
+ extra effort for difficult tasks and long-running asynchronous work.
20
+
9
21
  ## [0.7.0] - 2026-07-17
10
22
 
11
23
  ### Added
@@ -196,8 +208,9 @@ semantic version tags.
196
208
  - Initial shared routing table, cross-vendor dispatcher, runners, installer,
197
209
  and baseline lint fixes.
198
210
 
199
- [Unreleased]: https://github.com/Seraphim0916/omnilane/compare/v0.7.0...HEAD
211
+ [Unreleased]: https://github.com/Seraphim0916/omnilane/compare/v0.7.1...HEAD
200
212
 
213
+ [0.7.1]: https://github.com/Seraphim0916/omnilane/compare/v0.7.0...v0.7.1
201
214
  [0.7.0]: https://github.com/Seraphim0916/omnilane/compare/v0.6.0...v0.7.0
202
215
  [0.6.0]: https://github.com/Seraphim0916/omnilane/compare/v0.5.1...v0.6.0
203
216
  [0.5.1]: https://github.com/Seraphim0916/omnilane/compare/v0.5.0...v0.5.1
package/README.ja.md CHANGED
@@ -20,6 +20,16 @@
20
20
 
21
21
  ---
22
22
 
23
+ ## v0.7.1 の新機能
24
+
25
+ - **ルーティング表を更新(2026-07 モデルデータ)** — hardest-coding の第一候補を
26
+ GPT-5.6 Sol **max** に変更。Artificial Analysis Coding Agent Index v1.1 で
27
+ Sol (max) が 80 点の現行最高を記録し、旧「xhigh が max を上回る」スナップ
28
+ ショットを置き換えました。
29
+ - **Claude バックアップを強化** — hardest-coding と hard-judgment の
30
+ Claude Opus 4.8 バックアップを **xhigh** に変更。難しいタスクと長時間作業には
31
+ extra effort を推奨する Anthropic 公式ガイダンスに従います。
32
+
23
33
  ## v0.7.0 の新機能
24
34
 
25
35
  - **ディスパッチを事前にプレビュー** — `--dry-run` は解決済みの実行計画
@@ -108,10 +118,10 @@ flowchart LR
108
118
 
109
119
  | レーン | 第一候補 | バックアップ | 用途 |
110
120
  |---|---|---|---|
111
- | 🔥 hardest-coding | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 最難関の実装、根本原因デバッグ、正確性が要の変更 |
121
+ | 🔥 hardest-coding | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 最難関の実装、根本原因デバッグ、正確性が要の変更 |
112
122
  | 🏗️ bulk-mechanical | GPT-5.6 Terra (max) | Claude Sonnet 5 (high) | リファクタ、移行、テスト、大規模スイープ |
113
123
  | 🧹 triage | GPT-5.6 Luna (medium) | Gemini 3.5 Flash (Low) | 大量の一次スクリーニング |
114
- | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (high) | アーキテクチャ裁定、深い推論、セカンドオピニオン |
124
+ | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | アーキテクチャ裁定、深い推論、セカンドオピニオン |
115
125
  | ✒️ taste-final | Claude Opus 4.8 (high) | GPT-5.6 Sol (max) | 対外文章、prompt/ドキュメント推敲、スタイル最終審 |
116
126
  | 💬 consult | 明示指定したベンダー/モデル | —(フォールバックなし) | 自然言語で直接相談。`--vendor` を必ず維持 |
117
127
  | 🎨 ui-draft | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | デザインシステム/参考画像がある場合の UI ドラフト |
package/README.ko.md CHANGED
@@ -20,6 +20,16 @@
20
20
 
21
21
  ---
22
22
 
23
+ ## v0.7.1 새 기능
24
+
25
+ - **라우팅 테이블 갱신(2026-07 모델 데이터)** — hardest-coding 1순위를
26
+ GPT-5.6 Sol **max** 로 변경. Artificial Analysis Coding Agent Index v1.1 에서
27
+ Sol (max) 이 80점으로 현행 최고를 기록해, 기존 「xhigh 가 max 보다 낫다」
28
+ 스냅샷을 대체합니다.
29
+ - **Claude 백업 강화** — hardest-coding 과 hard-judgment 의 Claude Opus 4.8
30
+ 백업을 **xhigh** 로 변경. 어려운 작업과 장시간 작업에 extra effort 를
31
+ 권장하는 Anthropic 공식 가이드를 따릅니다.
32
+
23
33
  ## v0.7.0 새 기능
24
34
 
25
35
  - **디스패치를 먼저 미리보기** — `--dry-run` 은 완전히 해석된 실행 계획(vendor,
@@ -108,10 +118,10 @@ flowchart LR
108
118
 
109
119
  | 레인 | 1순위 모델 | 백업 | 용도 |
110
120
  |---|---|---|---|
111
- | 🔥 hardest-coding | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 가장 어려운 구현, 근본 원인 디버깅, 정확성이 핵심인 수정 |
121
+ | 🔥 hardest-coding | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 가장 어려운 구현, 근본 원인 디버깅, 정확성이 핵심인 수정 |
112
122
  | 🏗️ bulk-mechanical | GPT-5.6 Terra (max) | Claude Sonnet 5 (high) | 리팩터링, 마이그레이션, 테스트, 대량 스윕 |
113
123
  | 🧹 triage | GPT-5.6 Luna (medium) | Gemini 3.5 Flash (Low) | 대량 1차 선별 |
114
- | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (high) | 아키텍처 중재, 깊은 추론, 세컨드 오피니언 |
124
+ | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 아키텍처 중재, 깊은 추론, 세컨드 오피니언 |
115
125
  | ✒️ taste-final | Claude Opus 4.8 (high) | GPT-5.6 Sol (max) | 대외 문장, prompt/문서 다듬기, 스타일 최종심 |
116
126
  | 💬 consult | 명시적으로 지정한 벤더/모델 | —(폴백 없음) | 자연어 직접 상담. `--vendor` 를 반드시 유지 |
117
127
  | 🎨 ui-draft | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 디자인 시스템/참고 이미지가 있을 때의 UI 초안 |
package/README.md CHANGED
@@ -20,6 +20,17 @@ Every subtask goes to the model that is actually best at it — across<br/>
20
20
 
21
21
  ---
22
22
 
23
+ ## What's new in v0.7.1
24
+
25
+ - **Routing refresh (2026-07 model data)** — hardest-coding now dispatches
26
+ GPT-5.6 Sol at **max** effort: Artificial Analysis Coding Agent Index v1.1
27
+ scores Sol (max) at 80, the current state of the art, retiring the older
28
+ xhigh-beats-max snapshot.
29
+ - **Claude backups sharpened** — the Claude Opus 4.8 fallback on
30
+ hardest-coding and hard-judgment moves to **xhigh** effort, following
31
+ Anthropic's guidance to use extra effort for difficult tasks and
32
+ long-running work.
33
+
23
34
  ## What's new in v0.7.0
24
35
 
25
36
  - **Preview any dispatch first** — `--dry-run` prints the fully resolved plan
@@ -109,10 +120,10 @@ flowchart LR
109
120
 
110
121
  | Lane | First choice | Backup | When |
111
122
  |---|---|---|---|
112
- | 🔥 hardest-coding | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | Hardest implementation, deep root-cause debug, correctness-critical edits |
123
+ | 🔥 hardest-coding | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | Hardest implementation, deep root-cause debug, correctness-critical edits |
113
124
  | 🏗️ bulk-mechanical | GPT-5.6 Terra (max) | Claude Sonnet 5 (high) | Refactors, migrations, tests, review sweeps — mechanical endurance |
114
125
  | 🧹 triage | GPT-5.6 Luna (medium) | Gemini 3.5 Flash (Low) | High-volume scans, first-pass filtering |
115
- | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (high) | Architecture arbitration, deep reasoning, second opinions |
126
+ | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | Architecture arbitration, deep reasoning, second opinions |
116
127
  | ✒️ taste-final | Claude Opus 4.8 (high) | GPT-5.6 Sol (max) | User-facing prose, prompt/doc polish, style arbitration |
117
128
  | 💬 consult | Explicit named vendor/model | — (no fallback) | Direct natural-language consultation; always keep `--vendor` |
118
129
  | 🎨 ui-draft | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | UI drafts only WITH a design system / reference images |
package/README.zh-CN.md CHANGED
@@ -20,6 +20,14 @@
20
20
 
21
21
  ---
22
22
 
23
+ ## v0.7.1 新功能
24
+
25
+ - **路由表更新(2026-07 模型数据)** — hardest-coding 首选改为 GPT-5.6 Sol
26
+ **max** 档位:Artificial Analysis Coding Agent Index v1.1 测得 Sol (max)
27
+ 80 分为当前最高,替换旧的「xhigh 胜 max」快照。
28
+ - **Claude 备选升档** — hardest-coding 与 hard-judgment 的 Claude Opus 4.8
29
+ 备选改为 **xhigh**,依 Anthropic 官方对困难任务与长时间工作的建议。
30
+
23
31
  ## v0.7.0 新功能
24
32
 
25
33
  - **先预览再派工** — `--dry-run` 打印完整解析后的派工计划(vendor、模型、
@@ -99,10 +107,10 @@ flowchart LR
99
107
 
100
108
  | 通道 | 首选模型 | 备选模型 | 用途 |
101
109
  |---|---|---|---|
102
- | 🔥 hardest-coding | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 最难的实现、深度调试、正确性攸关的修改 |
110
+ | 🔥 hardest-coding | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 最难的实现、深度调试、正确性攸关的修改 |
103
111
  | 🏗️ bulk-mechanical | GPT-5.6 Terra (max) | Claude Sonnet 5 (high) | 重构、迁移、测试、大面积扫描——机械耐力活 |
104
112
  | 🧹 triage | GPT-5.6 Luna (medium) | Gemini 3.5 Flash (Low) | 高量初筛、第一轮过滤 |
105
- | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (high) | 架构仲裁、深度推理、第二意见 |
113
+ | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 架构仲裁、深度推理、第二意见 |
106
114
  | ✒️ taste-final | Claude Opus 4.8 (high) | GPT-5.6 Sol (max) | 对外文字、prompt 与文档打磨、风格终审 |
107
115
  | 💬 consult | 明确指定的厂商/模型 | —(不降级) | 自然语言直接咨询;必须保留 `--vendor` |
108
116
  | 🎨 ui-draft | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 有设计规范/参考图时的 UI 出稿;开放式视觉品味交给 taste-final |
package/README.zh-TW.md CHANGED
@@ -20,6 +20,14 @@
20
20
 
21
21
  ---
22
22
 
23
+ ## v0.7.1 新功能
24
+
25
+ - **路由表更新(2026-07 模型數據)** — hardest-coding 首選改為 GPT-5.6 Sol
26
+ **max** 檔位:Artificial Analysis Coding Agent Index v1.1 測得 Sol (max)
27
+ 80 分為現任最高,汰換舊的「xhigh 勝 max」快照。
28
+ - **Claude 備援升檔** — hardest-coding 與 hard-judgment 的 Claude Opus 4.8
29
+ 備援改為 **xhigh**,依 Anthropic 官方對困難任務與長時間工作的建議。
30
+
23
31
  ## v0.7.0 新功能
24
32
 
25
33
  - **先預覽再派工** — `--dry-run` 印出完整解析後的派工計畫(vendor、模型、
@@ -99,10 +107,10 @@ flowchart LR
99
107
 
100
108
  | 通道 | 首選模型 | 備選模型 | 用途 |
101
109
  |---|---|---|---|
102
- | 🔥 hardest-coding | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 最難的實作、深度除錯、正確性攸關的修改 |
110
+ | 🔥 hardest-coding | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 最難的實作、深度除錯、正確性攸關的修改 |
103
111
  | 🏗️ bulk-mechanical | GPT-5.6 Terra (max) | Claude Sonnet 5 (high) | 重構、搬遷、測試、大面積掃描——機械耐力活 |
104
112
  | 🧹 triage | GPT-5.6 Luna (medium) | Gemini 3.5 Flash (Low) | 高量初篩、第一輪過濾 |
105
- | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (high) | 架構仲裁、深度推理、第二意見 |
113
+ | ⚖️ hard-judgment | GPT-5.6 Sol (max) | Claude Opus 4.8 (xhigh) | 架構仲裁、深度推理、第二意見 |
106
114
  | ✒️ taste-final | Claude Opus 4.8 (high) | GPT-5.6 Sol (max) | 對外文字、prompt 與文件打磨、風格終審 |
107
115
  | 💬 consult | 明確點名的廠商/模型 | —(不降級) | 自然語言直接諮詢;必須保留 `--vendor` |
108
116
  | 🎨 ui-draft | GPT-5.6 Sol (xhigh) | Claude Opus 4.8 (high) | 有設計規範/參考圖時的 UI 出稿;開放式視覺品味交給 taste-final |
package/VERSION CHANGED
@@ -1 +1 @@
1
- 0.7.0
1
+ 0.7.1
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "omnilane",
3
- "version": "0.7.0",
3
+ "version": "0.7.1",
4
4
  "description": "One routing table, every harness — classify subtasks into lanes and dispatch each lane to the best vendor's agentic CLI (Codex, Claude, Gemini, Grok) using your existing subscription logins.",
5
5
  "bin": {
6
6
  "omnilane": "bin/omnilane"
package/routing.yaml CHANGED
@@ -12,10 +12,10 @@
12
12
  # standard short-context API tier — on subscription CLIs treat $ as relative ranking.
13
13
  # Your own job outcomes (~/.omnilane/jobs/) outrank these priors; edit lanes to match.
14
14
 
15
- hardest-coding: codex gpt-5.6-sol xhigh | claude claude-opus-4-8 high # AA coding 78.35; xhigh beats max (77.39) at lower cost — verified
15
+ hardest-coding: codex gpt-5.6-sol max | claude claude-opus-4-8 xhigh # AA Coding Agent Index v1.1: Sol (max) 80 = SOTA (2026-07-09, Codex harness); Claude xhigh per Anthropic guidance for difficult tasks
16
16
  bulk-mechanical: codex gpt-5.6-terra max | claude claude-sonnet-5 high | gemini "Gemini 3.5 Flash (High)" - # Terra 76.66 ~= Fable-class at a fraction of the cost
17
17
  triage: codex gpt-5.6-luna medium | gemini "Gemini 3.5 Flash (Low)" - | claude claude-haiku-4-5 - # high-volume scans
18
- hard-judgment: codex gpt-5.6-sol max | claude claude-opus-4-8 high # Sol max Intelligence 58.9 — second only to Fable 5 (59.9, fallback composite)
18
+ hard-judgment: codex gpt-5.6-sol max | claude claude-opus-4-8 xhigh # Sol max Intelligence 59 — second only to Fable 5 (2026-07-09 AA); Claude xhigh per Anthropic guidance for difficult tasks
19
19
  taste-final: claude claude-opus-4-8 high | codex gpt-5.6-sol max # user-facing prose, prompt/doc polish, Chinese phrasing, style arbitration
20
20
  consult: codex gpt-5.6-sol max | claude claude-opus-4-8 high | grok grok-4.5 - | gemini "Gemini 3.1 Pro (High)" - # direct named-model consultation; use --vendor to prevent fallback
21
21
  ui-draft: codex gpt-5.6-sol xhigh | claude claude-opus-4-8 high # only with a design system / reference images; open-ended visual taste -> taste-final