AI-IDE — Privacy Policy
Last updated: 2026-06-02
AI-IDE (“the App”) is a local code/Markdown editor that lets you collaborate with large-language-model (LLM) AI services that you choose and configure. This policy explains exactly what data the App handles, how it is collected, how it is used, and what is shared with third-party AI services.
Summary: The App does not operate any server of its own and does not collect analytics. The only outbound transmission of your content happens when you ask the AI to do something — at that point your message and the relevant file content are sent directly to the AI provider you configured. The App asks for your explicit consent before this happens for the first time.
1. Data the App handles
Stored only on your device
- API keys for the AI / web-search providers you add — stored in the iOS Keychain. Never written to the App’s database and never included in device backups.
- Chat history, provider settings, and app preferences — stored in a local on-device database.
- Your files — accessed through Apple’s Files (security-scoped bookmarks). The App does not copy your files to any server of its own.
- Imported files — when you import a file from Files, Photos, Camera, or the clipboard (the in-app “upload” action), it is copied into the workspace folder on your device only. Importing transmits nothing to any server. The imported file’s contents leave your device only if you later attach it, reference it with
@path, or ask the AI to edit it — at which point it is sent to the AI provider exactly as described below.
Sent to the AI service you choose
When — and only when — you send a chat message or ask the AI to read/edit a file, the following is transmitted to the AI provider endpoint you configured:
- The text you type in the chat box.
- The contents of files you reference with
@path or have open in the editor.
- Images you attach.
- When you ask the AI to edit files, the directory listing and file contents needed to perform that edit (always within the workspace you selected).
2. Who the data is sent to
The App does not route your data through any AI-IDE-owned proxy or server. Data is sent directly from your device to the LLM endpoint you configure under Settings → LLM Provider. This may be, for example:
- OpenAI (api.openai.com)
- Anthropic (api.anthropic.com)
- Google Gemini (generativelanguage.googleapis.com)
- Any other OpenAI-/Anthropic-compatible endpoint you enter, including self-hosted or local models.
Each of these third parties has its own privacy policy and data-handling practices, which govern what they do with the data once received. We encourage you to review the privacy policy of whichever provider you use. Optionally, if you enable web search, your search query is likewise sent directly to the search backend you configure (Brave, Tavily, or Perplexity).
3. Your consent
Before the App transmits any of your content to a third-party AI service for the first time, it presents a disclosure screen describing what will be sent and to whom, and asks you to explicitly agree. No data is sent until you accept. You can review your consent status, and withdraw it at any time, under Settings → Privacy & Data → AI Data-Sharing Consent. Withdrawing consent causes the disclosure to be shown again before the next transmission.
4. How the data is used
- Content sent to an AI provider is used solely to generate the AI’s response to your request (e.g. answering a question, proposing an edit).
- The App itself does not use your content for advertising, profiling, or any secondary purpose.
- The App collects no telemetry, usage analytics, or advertising identifiers.
5. Data we do NOT collect or share
- We do not collect personal data on any AI-IDE server (there is none).
- We do not upload your files to AI-IDE-owned infrastructure.
- We do not share your API keys with anyone; they remain in your device Keychain.
- We do not sell or rent any data.
6. Children
The App is not directed at children and does not knowingly collect data from children.
7. Changes
If this policy changes materially, the in-app consent version is incremented and you will be asked to review the updated disclosure before sending further data.
8. Contact
Questions about this policy: sportusagi@gmail.com
AI-IDE — 隱私權政策
最後更新:2026-06-02
AI-IDE(以下稱「本 App」)是一款在本機運作的程式碼 / Markdown 編輯器,讓你與自行選擇並設定的大型語言模型(LLM)AI 服務協作。本政策說明本 App 處理哪些資料、如何蒐集、如何使用,以及哪些內容會分享給第三方 AI 服務。
摘要:本 App 不營運任何自有伺服器,也不蒐集分析資料。唯一會將你的內容送出的時機,是你主動要求 AI 處理時——此時你的訊息與相關檔案內容會直接傳送到你所設定的 AI 服務商。首次傳送前,本 App 會先取得你的明確同意。
1. 本 App 處理的資料
僅儲存在你的裝置上
- API Key(你新增的 AI / 網路搜尋服務商)——儲存於 iOS Keychain,絕不寫入 App 資料庫,也不隨裝置備份。
- 對話紀錄、服務商設定與 App 偏好——儲存於本機資料庫。
- 你的檔案——透過 Apple 檔案授權(security-scoped bookmark)存取,本 App 不會將檔案複製到任何自有伺服器。
- 匯入的檔案——當你從「檔案」、相簿、相機或剪貼簿匯入檔案(App 內的「上傳」動作)時,檔案只會被複製到你裝置上的工作目錄,匯入過程不會傳送任何資料到任何伺服器。匯入的檔案內容唯有在你之後將其附加、以
@path 引用、或要求 AI 編輯時才會離開你的裝置——此時才會依下文所述傳送給 AI 服務商。
傳送給你所選擇的 AI 服務
只有在你送出對話訊息、或要求 AI 讀取 / 編輯檔案時,以下內容才會傳送到你設定的 AI 服務商端點:
- 你在對話框輸入的文字。
- 你以
@path 引用或正在編輯器中開啟的檔案內容。
- 你附加的圖片。
- 當你要求 AI 編輯檔案時,執行該編輯所需的目錄清單與檔案內容(一律限於你所選擇的工作目錄內)。
2. 資料傳送對象
本 App 不會透過任何 AI-IDE 自有的代理或伺服器轉送你的資料。資料是從你的裝置直接送往你在「設定 → LLM Provider」所設定的 LLM 端點,例如:
- OpenAI(api.openai.com)
- Anthropic(api.anthropic.com)
- Google Gemini(generativelanguage.googleapis.com)
- 你自行輸入的其他 OpenAI / Anthropic 相容端點,包含自架或本地模型。
這些第三方各有自己的隱私政策與資料處理方式,資料送達後即由其政策規範。建議你查閱所使用服務商的隱私政策。若你啟用網路搜尋,搜尋字串同樣會直接送往你設定的搜尋後端(Brave、Tavily 或 Perplexity)。
3. 你的同意
本 App 首次將你的內容傳送給第三方 AI 服務前,會顯示說明畫面,告知將傳送哪些內容、傳送給誰,並請你明確同意。在你同意之前不會送出任何資料。你可隨時於「設定 → 隱私與資料 → AI 資料分享同意」檢視同意狀態並撤回同意;撤回後,下次傳送前會再次顯示說明。
4. 資料用途
- 傳送給 AI 服務商的內容,僅用於針對你的要求產生 AI 回應(例如回答問題、提出修改建議)。
- 本 App 本身不會將你的內容用於廣告、剖析或任何次要用途。
- 本 App 不蒐集遙測、使用分析或廣告識別碼。
5. 我們不會蒐集或分享的資料
- 不會在任何 AI-IDE 伺服器蒐集個人資料(本來就沒有伺服器)。
- 不會將你的檔案上傳到 AI-IDE 自有設施。
- 不會將你的 API Key 分享給任何人;它只留在你裝置的 Keychain。
- 不會販售或出租任何資料。
6. 兒童
本 App 並非以兒童為對象,也不會在知情下蒐集兒童資料。
7. 變更
若本政策有重大變更,App 內的同意版本號會遞增,並在你繼續傳送資料前請你重新檢視更新後的說明。
8. 聯絡方式
關於本政策的問題:sportusagi@gmail.com