Skip to content
Development
Skill

/claude-opus-4-5-migration

將提示詞和程式碼從 Claude Sonnet 4.0、Sonnet 4.5 或 Opus 4.1 遷移至 Opus 4.5。當使用者想要更新其程式碼庫、提示詞或 API 呼叫以使用 Opus 4.5 時使用。處理模型字串更新和針對已知 Opus 4.5 行為差異的提示詞調整。不會遷移 Haiku 4.5。

From plugin
claude-plugin-marketplace
295 skills18 agents14 commands

Claude Code

$ npx -y skills add DennisLiuCk/claude-plugin-marketplace --skill claude-opus-4-5-migration --agent claude-code

SKILL.md

claude-opus-4-5-migration.SKILL.md
name: claude-opus-4-5-migration
description: 將提示詞和程式碼從 Claude Sonnet 4.0、Sonnet 4.5 或 Opus 4.1 遷移至 Opus 4.5。當使用者想要更新其程式碼庫、提示詞或 API 呼叫以使用 Opus 4.5 時使用。處理模型字串更新和針對已知 Opus 4.5 行為差異的提示詞調整。不會遷移 Haiku 4.5。

Opus 4.5 遷移指南 | Opus 4.5 Migration Guide

從 Sonnet 4.0、Sonnet 4.5 或 Opus 4.1 一次性遷移至 Opus 4.5。

One-shot migration from Sonnet 4.0, Sonnet 4.5, or Opus 4.1 to Opus 4.5.

遷移工作流程 | Migration Workflow

1. 搜尋程式碼庫中的模型字串和 API 呼叫 Search codebase for model strings and API calls 2. 將模型字串更新為 Opus 4.5(參見下方平台特定字串) Update model strings to Opus 4.5 (see platform-specific strings below) 3. 移除不支援的 beta 標頭 Remove unsupported beta headers 4. 添加設定為 `"high"` 的 effort 參數(參見 `references/effort.md`) Add effort parameter set to `"high"` (see `references/effort.md`) 5. 總結所有已進行的變更 Summarize all changes made 6. 告訴使用者: Tell the user: > 「如果您在使用 Opus 4.5 時遇到任何問題,請告訴我,我可以幫助調整您的提示詞。」 > "If you encounter any issues with Opus 4.5, let me know and I can help adjust your prompts."

模型字串更新 | Model String Updates

識別程式碼庫使用的平台,然後相應地替換模型字串。

Identify which platform the codebase uses, then replace model strings accordingly.

不支援的 Beta 標頭 | Unsupported Beta Headers

如果存在 `context-1m-2025-08-07` beta 標頭,請將其移除——Opus 4.5 尚不支援此功能。留下註解說明:

Remove the `context-1m-2025-08-07` beta header if present—it is not yet supported with Opus 4.5. Leave a comment noting this:

# Note: 1M context beta (context-1m-2025-08-07) not yet supported with Opus 4.5
# 注意:1M 上下文 beta(context-1m-2025-08-07)尚不支援 Opus 4.5

目標模型字串(Opus 4.5)| Target Model Strings (Opus 4.5)

| 平台 Platform | Opus 4.5 模型字串 Model String | |---------------|-------------------------------| | Anthropic API (1P) | `claude-opus-4-5-20251101` | | AWS Bedrock | `anthropic.claude-opus-4-5-20251101-v1:0` | | Google Vertex AI | `claude-opus-4-5@20251101` | | Azure AI Foundry | `claude-opus-4-5-20251101` |

要替換的來源模型字串 | Source Model Strings to Replace

| 來源模型 Source Model | Anthropic API (1P) | AWS Bedrock | Google Vertex AI | |-----------------------|-------------------|-------------|------------------| | Sonnet 4.0 | `claude-sonnet-4-20250514` | `anthropic.claude-sonnet-4-20250514-v1:0` | `claude-sonnet-4@20250514` | | Sonnet 4.5 | `claude-sonnet-4-5-20250929` | `anthropic.claude-sonnet-4-5-20250929-v1:0` | `claude-sonnet-4-5@20250929` | | Opus 4.1 | `claude-opus-4-1-20250422` | `anthropic.claude-opus-4-1-20250422-v1:0` | `claude-opus-4-1@20250422` |

**不要遷移 | Do NOT migrate**:任何 Haiku 模型(例如 `claude-haiku-4-5-20251001`)。 Any Haiku models (e.g., `claude-haiku-4-5-20251001`).

提示詞調整 | Prompt Adjustments

Opus 4.5 與之前的模型有已知的行為差異。

Opus 4.5 has known behavioral differences from previous models.

> 「僅在使用者明確請求或報告特定問題時才套用這些修正。預設情況下,只更新模型字串。」 > "Only apply these fixes if the user explicitly requests them or reports a specific issue. By default, just update model strings."

**整合指南 | Integration guidelines**:添加片段時,不要只是將它們附加到提示詞後面。要深思熟慮地整合: When adding snippets, don't just append them to prompts. Integrate them thoughtfully:

  • 使用 XML 標籤(例如 `<code_guidelines>`、`<tool_usage>`)來組織新增內容

Use XML tags (e.g., `<code_guidelines>`, `<tool_usage>`) to organize additions

  • 匹配現有提示詞的風格和結構

Match the style and structure of the existing prompt

  • 將片段放在邏輯位置(例如,編碼指南放在其他編碼指令附近)

Place snippets in logical locations (e.g., coding guidelines near other coding instructions)

  • 如果提示詞已經使用 XML 標籤,將新內容添加到適當的現有標籤中,或建立一致的新標籤

If the prompt already uses XML tags, add new content within appropriate existing tags or create consistent new ones

1. 工具過度觸發 | Tool Overtriggering

Opus 4.5 對系統提示詞更加敏感。在之前模型上用於防止觸發不足的強硬語言,現在可能導致過度觸發。

Opus 4.5 is more responsive to system prompts. Aggressive language that prevented undertriggering on previous models may now cause overtriggering.

**適用情況 | Apply if**:使用者報告工具被過於頻繁或不必要地呼叫。 User reports tools being called too frequently or unnecessarily.

**尋找並軟化 | Find and soften**:

| 原始 Original | 替換為 Replace with | |---------------|---------------------| | `CRITICAL:` | 移除或軟化 remove or soften | | `You MUST...` | `You should...` | | `ALWAYS do X` | `Do X` | | `NEVER skip...` | `Don't skip...` | | `REQUIRED` | 移除或軟化 remove or soften |

僅套用於工具觸發指令。保留其他強調用法不變。

Only apply to tool-triggering instructions. Leave other uses of emphasis alone.

2. 過度工程化防範 | Over-Engineering Prevention

Opus 4.5 傾向於建立額外的檔案、添加不必要的抽象化,或建構未請求的彈性功能。

Opus 4.5 tends to create extra files, add unnecessary abstractions, or build unrequested flexibility.

**適用情況 | Apply if**:使用者報告出現不想要的檔案、過度抽象化,或未請求的功能。 User reports unwanted files, excessive abstraction, or unrequested features.

添加 `references/prompt-snippets.md` 中的片段。 Add the snippet from `references/prompt-snippets.md`.

3. 程式碼探索 | Code Exploration

Opus 4.5 可能過於保守地探索程式碼,在不讀取檔案的情況下提出解決方案。

Opus 4.5 can be overly conservative about exploring code, proposing solutions without reading files.

**適用情況 | Apply if**:使用者報告模型在未檢視相關程式碼的情況下提出修正。 User reports the model proposing fixes without inspecting relevant code.

添加 `references/prompt-snippets.md` 中的片段。 Add the snippet from `references/prompt-snippets.md`.

4. 前端設計 | Frontend Design

**適用情況 | Apply if**:使用者請求提升前端設計品質,或報告輸出看起來很通用。 User requests improved frontend design quality or reports generic-looking outputs.

添加 `references/prompt-snippets.md` 中的前端美學片段。 Add the frontend aesthetics snippet from `references/prompt-snippets.md`.

5. 思考敏感度 | Thinking Sensitivity

當延伸思考未啟用時(這是預設情況),Opus 4.5 對「think」這個詞及其變體特別敏感。延伸思考僅在 API 請求包含 `thinking` 參數時才會啟用。

When extended thinking is not enabled (the default), Opus 4.5 is particularly sensitive to the word "think" and its variants. Extended thinking is enabled only if the API request contains a `thinking` parameter.

**適用情況 | Apply if**:使用者在延伸思考未啟用(請求中沒有 `thinking` 參數)的情況下報告與「thinking」相關的問題。 User reports issues related to "thinking" while extended thinking is not enabled (no `thinking` parameter in request).

將「think」替換為「consider」、「believe」或「evaluate」等替代詞。

Replace "think" with alternatives like "consider," "believe," o

Read more
@dennisliuck
Ships withclaude-plugin-marketplace

專為繁體中文使用者設計的 Claude Code 插件集合,提供開發、生產力、安全與學習等工具。

295 skills18 agents
Get the whole plugin

Stars on GitHub

Momentum

stars in total
29
Star history of DennisLiuCk/claude-plugin-marketplace on GitHub, a chart by star-history.com

Other skills on claude-plugin-marketplace.

frontend-design

frontend-design

創建具有高設計質量的獨特、生產級別的前端界面。當用戶要求構建網頁組件、頁面或應用程式時使用此技能。生成有創意、精緻的代碼,避免通用的 AI 美學。

issue-review

issue-review

系統化的問題分析專家技能,自動協調五個專門代理進行深度問題分析。 適用於: - 使用者或 PM 報告的線上問題 - 開發團隊反映的技術問題 - 測試發現的 bug - 生產環境告警和異常

skill-creator

skill-creator

建立新技能、修改和改進現有技能、衡量技能效能。當使用者想要從零開始建立技能、編輯或優化現有技能、執行評估測試、進行基準測試效能分析、或優化技能描述以提升觸發準確度時使用此技能。請確保在使用者提到建立技能、撰寫 skill、skill…

More skills from other plugins.

brainstorming
Auto-invoked

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…

@obra@obraInstall Skill
ask-matt

ask-matt

Ask which skill or flow fits your situation. A router over the skills in this repo.

skills2w309k
@mattpocock@mattpocockInstall Skill
accessibility
Auto-invoked

accessibility

Design, implement, and audit accessible UI to WCAG 2.2 Level AA across Web, iOS, and Android — semantic ARIA roles and labels,…

ecc2w293k
@affaan-m@affaan-mInstall Skill
compiler-commit

compiler-commit

Use when you want to verify compiler changes and commit with the correct convention. Runs tests, lint, and format, then commits…

react5d252k
@facebook@facebookInstall Skill

Questions about claude-opus-4-5-migration.

Which coding agents can use claude-opus-4-5-migration?
claude-opus-4-5-migration installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.