请先阅读：Agentale是一款 AI 编程与创作工具。部分开发、构建、调试、审核和发布工作可能在服务器侧完成，因此完成你主动发起的在线任务时，需要回传提示词、代码、文件、日志或其他必要信息。请勿把密码、API 密钥、私钥或无权处理的个人信息直接写入提示词、代码或公开作品。
1. 适用范围与运营者
本政策适用于 Agentale Android 与 iOS 应用、官网及相关云服务。个人信息处理者和服务运营者为海南福驰科技有限公司（以下称“我们”）。隐私与客服联系邮箱：fuchitech@qq.com。
本政策不替代你主动选择的第三方 AI 模型、代码托管、对象存储或外部工作流服务的隐私政策。使用这些第三方服务前，请同时阅读其条款。
2. 我们可能处理的信息
- 账户与身份信息：账号、显示名称、邮箱或其他登录标识、密码的不可逆校验值、会话状态、恢复状态和设备安装标识。我们不应保存可还原的明文密码。
- 创作与开发信息：你输入的需求、提示词、对话、项目名称、源代码、配置、文件、图片、音频、视频、文档、版本、测试结果、构建产物、封面、说明和作品关系。
- 服务器开发与调试信息：任务指令、依赖信息、命令及结果、编译和测试日志、错误堆栈、运行状态、资源用量、审核结果以及为定位问题所需的上下文。
- 发布与社区信息：作品码、公开链接、作者展示名、版本、年龄分级、分享设置、二创许可、收藏、举报、申诉和审核记录。
- 设备与服务记录：应用版本、操作系统、语言、网络状态、IP 地址、访问时间、崩溃与诊断日志、安全事件及必要审计记录。IP 可能推断出大致地区，但我们不将其用于精确定位。
- 订阅信息：商品、应用商店、订单或交易标识、购买状态和权益状态。完整银行卡号由 Apple App Store、Google Play 等支付服务商处理，我们不接收。
- 客服与反馈：你主动提供的联系方式、问题描述、附件、沟通和处理记录。
3. 服务器侧开发、调试与代码保管
为完成你主动发起的 AI 编程、云端构建、依赖安装、自动测试、错误诊断、版本审核、发布或恢复任务，客户端会把完成任务所必需的信息发送到服务器。这些信息可能包括需求、对话、源代码、选定文件、资源清单、构建配置、错误日志和测试结果。我们仅应在实现相应功能、安全保障和故障处理所需范围内处理。
如果你选择云同步、云端项目、发布、分享、联合发行或其他需要服务器保存的功能，代码、文件、版本和相关记录可能保存在服务器或受托服务商的系统中。仅使用本地功能时，本地工作区不会因安装应用而自动公开；但你调用在线 AI 或云端开发功能时，相应请求所需内容仍会传输。
请使用环境变量、平台密钥管理或专门的秘密存储功能保管密码、API 密钥、访问令牌、私钥和生产凭据，不要硬编码进项目、粘贴进对话或加入发布包。因你未妥善保管、主动公开、错误授权或把秘密写入可分享内容，且并非由我们的过错造成的损失，由你依法承担相应责任；本条不排除我们依法应承担的数据安全、隐私保护或其他不可免责责任。
4. 处理目的与依据
我们为提供你请求的服务而创建和维护账户、执行开发和调试任务、保存项目、同步版本、运行测试、生成预览、审核和发布作品、生成分享链接、恢复购买以及提供客服。我们还会处理必要日志，以保护账户、阻止滥用、保障系统稳定、排查故障、履行法律义务和处理投诉。
创建分享链接时，服务器会把你主动提交的冻结版本中与判断有关的最少内容（例如名称、说明、创作要求、代码或文本摘要、权限清单、年龄分级和首张预览图）交给受托 AI 模型进行内容安全判断，用于识别未成年人伤害、暴力、仇恨歧视、隐私泄露及内容规范列明的其他严重风险。结论不明确或处理失败达到重试上限时，作品保持私密并进入人工复核。
需要你同意的可选统计、模型改进数据或非必要权限将单独说明并允许撤回。产品使用统计同意不等于允许将完整项目用于模型训练；上述分享安全判断也不等于允许将作品用于通用模型训练。除非另行显著说明并取得所需授权，我们不会把你的私有代码或作品用于训练通用模型。
5. 本机数据与设备权限
本地项目、聊天、任务、模型设置和运行记录通常保存在设备或你选择的工作目录。清理缓存、卸载应用、删除工作目录或设备损坏可能导致本地数据丢失，请自行保留必要备份。
照片/文件、相机、麦克风和通知权限只在对应功能需要时按场景请求。例如，你选择图片、拍摄素材、录制声音或接收任务通知时才需要相应能力。拒绝非核心权限不会阻止无关功能；你可以在系统设置中撤回。Agentale不因核心功能而要求读取联系人或精确位置。
6. 服务商、共享与披露
为提供服务，我们可能向受合同和保密义务约束的服务商提供最少必要信息，包括云计算、数据库、对象存储/CDN、安全监控、客服、身份认证、Apple App Store 与 Google Play 购买验证，以及你选择或配置的 AI 模型服务商。相关服务商仅应按约定目的处理数据。
在法律要求、保护人员和系统安全、调查欺诈或攻击、处理权利投诉或企业重组所必要时，我们也可能依法披露信息。我们不出售个人信息，也不将个人信息用于跨应用行为广告。
7. 公开发布、分享与二创
只有在你主动选择并确认发布或分享时，作品才会进入相应可见范围。“持链接可见”意味着获得链接的人可能访问并再次转发；公开展示可能让所有访问者看到作品、作者展示名、说明、封面、版本、作品码和你选择公开的素材。关闭链接可以阻止平台继续提供访问，但无法收回他人此前已合法保存的副本。
分享前请移除密码、密钥、个人证件、联系方式、生产数据库信息、未公开商业秘密和他人隐私。开启二创只授权服务范围内对已公开作品的相应使用，不会自动授权第三方私密素材，也不会替你取得原权利人的授权。
8. 保存期限与安全
账户和云端作品通常保存至你删除相应内容或账号。经确认的删除请求完成后，关联数据原则上在 30 日内从活动系统删除或去标识化，轮换备份原则上在 90 日内失效。安全、反滥用和发布审核记录通常最多保存 180 日；客服与申诉记录通常最多保存 24 个月；交易、税务、知识产权投诉或依法保全的数据按法定期限保存。
我们采用传输加密、访问控制、最小权限、凭据隔离、签名验证、审计、备份与安全修复等合理措施。任何系统都不能保证绝对安全；发生依法需要通知的数据安全事件时，我们将按法律要求通知受影响用户和主管机关。
9. 内容安全、禁止生成与禁止分享
你不得要求Agentale生成、保存、发布、分享或协助传播下列内容。平台可以在提示输入、AI 生成、构建、发布、生成链接、公开展示和举报处理阶段使用自动规则、模型安全能力及人工复核；疑似违规内容可能被拒绝生成、阻止分享、限制访问、下架、删除或进入复核。严重情形可能导致账号限制、证据保全或依法向主管机关报告。
- 色情与性内容露骨色情、性服务、性剥削、偷拍色情及涉及未成年人的性内容。
- 暴力与血腥虐杀、肢解、酷刑、真实血腥画面，或美化、鼓励暴力。
- 仇恨与歧视针对受保护群体的侮辱、仇恨或暴力煽动。
- 极端主义与恐怖主义宣传、招募、赞美或协助恐怖主义和极端主义。
- 毒品及受管制物品毒品制造、交易或违法获取处方药等。
- 武器与危险品违法购买、制造武器、爆炸物、危险化学品或具体伤害指导。
- 自杀、自残与危险行为鼓励或指导自杀、自残及极端危险挑战。
- 违法犯罪盗窃、诈骗、洗钱、绑架、勒索、制假、逃避执法等实质性协助。
- 赌博及博彩非法赌博、博彩引流或未获许可的受监管博彩服务。
- 欺诈与金融犯罪钓鱼、信用卡欺诈、虚假投资、资金盘或冒充诈骗。
- 网络攻击与恶意软件盗号、木马、勒索软件、窃取凭证或破坏他人系统。
- 隐私与个人信息侵犯人肉搜索、泄露敏感个人信息或非法跟踪监控。
- 骚扰、霸凌与威胁持续骚扰、恐吓、羞辱、勒索或现实伤害威胁。
- 未成年人不当内容诱骗、性化未成年人、儿童剥削及其他严重危害儿童的内容。
- 虚假身份与冒充以欺骗为目的冒充真人、伪造身份材料或制作 Deepfake。
- 知识产权侵权盗版、破解、绕过付费机制、仿冒品牌或大规模侵权。
- 垃圾信息与恶意营销批量骚扰邮件、垃圾短信、刷评论、刷量或虚假评价。
- 高风险误导信息把未经验证的医疗、金融、法律等信息包装成确定性专业结论。
- 动物虐待鼓励、展示或指导虐待动物。
- 其他违法或严重有害内容根据用户所在地法律、应用商店政策和实际风险处理的其他内容。
你应对自己提交、生成后选择使用、发布、分享、传播或授权二创的内容进行必要审查，并对自己的违法、侵权或违反规则的行为依法承担相应责任。平台依法负有的内容治理、隐私、安全和消费者保护义务不因本条免除。完整规则见《内容规范》。
10. 素材与知识产权责任
你只能上传、处理、生成、复制、发布和传播自己拥有合法权利，或已取得充分授权的代码、文字、图片、音频、视频、字体、数据、商标、肖像、姓名和其他素材。不得盗版、破解、仿冒品牌、未经授权复制作品，或侵犯著作权、商标权、专利权、肖像权、姓名权、名誉权、隐私权、商业秘密等合法权益。
你应保留许可证、购买凭证、授权书、开源声明和来源记录。因你提供或传播的内容侵犯第三方权利而产生的投诉、索赔、处罚或损失，由责任方依法承担；如给我们或第三方造成损失，你应在法律允许范围内承担相应赔偿责任。我们可以要求权利证明，并可对疑似侵权内容拒绝生成或分享、限制访问、移除或保全证据。
11. 账号、密码、密钥与用户责任
请妥善保管账号密码、验证码、恢复码、API 密钥、访问令牌、SSH 密钥、签名证书、云服务凭据及其他秘密。不要与他人共用账号，不要把秘密放入公开仓库、截图、日志、提示词或可分享作品，并及时撤销疑似泄露的凭据。
你应确保自己有权访问所连接的系统、仓库、域名、账户和数据，并对账号下由你授权的开发、执行、发布与传播行为负责。发现未经授权访问或秘密泄露时，请立即更换凭据并联系我们。因你违反上述义务且并非由我们的过错造成的损失，由你依法承担；我们对自身违法、违约、故意、重大过失或依法不得排除的责任继续负责。
12. 你的权利、选择与账号删除
根据适用法律，你可以请求访问、复制、导出、更正、删除或限制处理个人信息，撤回同意、反对特定处理、关闭可选统计，并向主管机关投诉。为保护账户，我们可能要求验证身份。撤回非必要授权不影响此前合法处理；删除服务必需信息可能导致相应云功能无法继续。
你可以在Agentale应用的账户与隐私设置中发起删除整个云账号，也可按账号删除说明通过 fuchitech@qq.com 提交请求。账号删除会注销会话、使平台控制的公开链接失效，并按保存期限删除或去标识化云数据；不会自动删除你设备上的本地工作区。自动续订需同时在相应的 Apple App Store 或 Google Play 订阅管理中取消。
13. 未成年人
Agentale不以 13 周岁以下儿童为目标用户，也不申请儿童类别。未满 18 周岁的用户应在监护人指导下阅读本政策；低于所在地独立同意数字服务年龄的用户不得自行创建云账号。若发现未经适当同意收集儿童个人信息，我们会依法尽快删除。涉及儿童性剥削或现实安全风险的内容将优先阻止、处置和升级。
14. 跨境处理
服务器、AI 模型或其他服务商可能位于你所在国家或地区之外。发生跨境处理时，我们将依据适用法律采用所需的安全评估、认证、标准合同、充分性决定或其他合法机制，并实施数据最小化、传输加密和访问控制。你自行配置境外服务商时，还应查看该服务商的数据处理地区与政策。
15. 联系我们与政策更新
我们可能随功能、法律、服务商或数据处理方式变化更新本政策。重大变化将通过应用、网站或其他显著方式通知，并在法律要求时重新取得同意。页面顶部会标明最新版本和生效日期。
Please read first: Agentale is an AI coding and creation tool. Some development, build, debugging, review, and publication tasks may run on servers. Completing an online task you initiate may therefore require sending prompts, code, files, logs, or other necessary information. Do not place passwords, API keys, private keys, or personal data you have no right to process in prompts, code, or public creations.
1. Scope and operator
This Policy applies to the Agentale Android and iOS apps, website, and related cloud services. The service operator and data controller is Hainan Fuchi Technology Co., Ltd. (“we”). Privacy and support contact: fuchitech@qq.com.
Third-party AI models, repositories, storage, and external workflows you choose may process data under their own policies.
2. Information we may process
- Account data: account and display names, email or login identifier, non-reversible password verifier, sessions, recovery state, and installation identifier.
- Creation data: requirements, prompts, conversations, project names, source code, configuration, files, media, documents, versions, tests, builds, covers, descriptions, and creation relationships.
- Server-side work: task instructions, dependencies, command results, build and test logs, errors, runtime state, resource usage, and review records needed to complete or diagnose a task.
- Publishing and community data: creation codes, links, display name, version, age rating, sharing and remix settings, favorites, reports, appeals, and moderation records.
- Device and service records: app version, operating system, language, network state, IP address, access time, crashes, diagnostics, security events, and required audit logs. IP may reveal an approximate area but is not used for precise location.
- Purchases: product, app store, transaction identifier, purchase state, and entitlement. Apple App Store, Google Play, or another processor handles full payment-card details.
- Support: contact details, descriptions, attachments, and correspondence you submit.
3. Server-side development, debugging, and code storage
To perform AI coding, cloud builds, dependency installation, automated tests, diagnostics, version review, publication, or recovery that you request, Agentale may send the minimum necessary prompts, conversations, source code, selected files, asset manifests, build configuration, errors, and test results to servers.
If you enable cloud projects, sync, publication, sharing, co-publishing, or another hosted feature, code, files, versions, and related records may remain on our or our processors’ systems. A local workspace is not made public automatically, but content needed for an online AI or development request is transmitted.
Use environment variables, platform key management, or dedicated secret storage for passwords, API keys, tokens, private keys, and production credentials. Do not hard-code or paste secrets into chats or published bundles. To the extent permitted by law, you are responsible for loss caused by your disclosure, misconfiguration, or unauthorized material where that loss is not attributable to us. This does not exclude our non-waivable privacy, security, or other legal duties.
4. Purposes and legal bases
We process data to provide the service you request: accounts, development and debugging, storage, sync, tests, previews, review, publishing, links, purchase restoration, and support. Necessary logs also protect accounts, prevent abuse, keep systems reliable, diagnose faults, comply with law, and resolve complaints.
Optional analytics, model-improvement data, or non-essential permissions require separate notice and may be withdrawn. Analytics consent is not consent to train on complete projects. We do not use private code or creations to train a general model unless we separately explain the use and obtain required permission.
5. On-device data and permissions
Local projects, conversations, tasks, model settings, and runtime history generally stay on your device or in a workspace you choose. Cache clearing, uninstalling, deleting the workspace, or device failure can remove local data; keep appropriate backups.
Photos/files, camera, microphone, and notifications are requested in context when a feature needs them. Denying a non-essential permission does not block unrelated features, and permissions can be withdrawn in system settings. Core Agentale features do not require contacts or precise location.
6. Processors, sharing, and disclosures
We disclose only the minimum necessary data to contracted processors such as cloud hosting, databases, object storage/CDNs, security monitoring, support, identity, Apple App Store and Google Play purchase verification, and AI-model providers you select or configure. We may also disclose information as required by law, to protect people or systems, investigate abuse, resolve rights claims, or support a corporate transaction. We do not sell personal data or use it for cross-app behavioral advertising.
7. Public publishing, sharing, and remixing
A creation enters the visibility you select only after you initiate and confirm publication or sharing. Anyone with a link may access and forward a link-visible creation. Public visitors may see the creation, display name, description, cover, version, code, and assets you publish. Disabling a link stops access controlled by us but cannot recall copies already saved lawfully.
Remove credentials, identity documents, contact details, production database information, confidential business material, and other people’s private data before sharing. Remix permission does not grant rights in third-party material beyond your lawful authority.
8. Retention and security
Accounts and hosted creations are normally kept until deleted. After a confirmed deletion request, associated data is generally deleted or de-identified from active systems within 30 days and expires from rotating backups within 90 days. Security, anti-abuse, and publication review records are generally retained up to 180 days; support and appeal records up to 24 months. Transactions, tax, rights claims, and legal holds follow applicable retention law.
We use reasonable measures such as encryption in transit, access control, least privilege, credential separation, signature verification, audits, backups, and security remediation. No system is absolutely secure. We will provide legally required breach notices.
9. Content safety; blocked generation and sharing
You must not use Agentale to generate, store, publish, share, or facilitate unlawful, infringing, or seriously harmful content. Automated rules, model safeguards, and human review may operate at prompt, generation, build, publication, link creation, public listing, or report stages. Suspected violations may be blocked, restricted, removed, reviewed, preserved as evidence, or reported as required by law.
- Sexual contentExplicit pornography, sexual services or exploitation, voyeuristic sexual material, and any sexual content involving minors.
- Graphic violenceKilling, dismemberment, torture, real gore, or glorification and encouragement of violence.
- Hate and discriminationAbuse or incitement against protected groups.
- Terrorism and extremismPromotion, recruitment, praise, or operational assistance.
- Drugs and controlled goodsManufacture, trade, or unlawful access.
- Weapons and dangerous materialsIllegal acquisition or manufacture and actionable harm instructions.
- Suicide and self-harmEncouragement, instruction, or extreme dangerous challenges.
- CrimeMeaningful assistance for theft, fraud, laundering, kidnapping, extortion, counterfeiting, or evading law enforcement.
- GamblingIllegal gambling, prohibited promotion, or unlicensed regulated services.
- Financial fraudPhishing, card fraud, fake investments, Ponzi schemes, or impersonation scams.
- Cyberattacks and malwareCredential theft, trojans, ransomware, or system destruction.
- Privacy abuseDoxxing, disclosure of sensitive personal data, or unlawful tracking.
- Harassment and threatsTargeted abuse, intimidation, humiliation, extortion, or real-world threats.
- Child endangermentGrooming, sexualization, exploitation, or other severe harm to children.
- Deceptive impersonationFraudulent identity documents, impersonation, or deceptive deepfakes.
- Intellectual-property infringementPiracy, cracking, paywall bypass, counterfeiting, or mass infringement.
- Spam and manipulationBulk harassment, spam messages, fake reviews, or metric manipulation.
- High-risk misinformationUnverified medical, financial, or legal claims presented as certain professional conclusions.
- Animal abuseEncouraging, depicting for gratification, or instructing animal cruelty.
- Other unlawful or severe harmOther material restricted by applicable law, store policy, or serious risk.
You must review content you submit or choose to use, publish, or share, and are legally responsible for your unlawful or infringing conduct. This does not waive duties imposed directly on us. See the Content Standards.
10. Assets and intellectual property
Use only code, text, images, audio, video, fonts, data, trademarks, likenesses, names, and other material you own or are properly authorized to process and distribute. Keep licenses, receipts, releases, open-source notices, and provenance records. You are responsible to the lawful extent for third-party claims caused by material you provide or distribute. We may request proof of rights and block or remove suspected infringement.
11. Accounts, passwords, keys, and your responsibility
Protect passwords, verification codes, recovery codes, API keys, tokens, SSH keys, signing certificates, cloud credentials, and other secrets. Do not share accounts or place secrets in public repositories, screenshots, logs, prompts, or shared creations. Rotate suspected compromised credentials promptly.
Use only systems, repositories, domains, accounts, and data you are authorized to access. You are responsible for development, execution, publication, and distribution that you authorize through your account. To the extent permitted by law, you bear loss caused by your breach of these duties where it is not attributable to us; we remain responsible for our own unlawful conduct, breach, willful misconduct, gross negligence, and non-waivable duties.
12. Your rights, choices, and account deletion
Depending on applicable law, you may request access, copies, export, correction, deletion, or restriction; withdraw consent; object to certain processing; disable optional analytics; and complain to a regulator. Identity verification may be required. Deleting necessary data may make a hosted feature unavailable.
Initiate cloud-account deletion in Agentale’s account and privacy settings or follow the account-deletion instructions to email fuchitech@qq.com. Deletion signs out sessions, disables links controlled by us, and deletes or de-identifies hosted data under the retention rules. It does not erase local workspaces. Cancel auto-renewal separately in the applicable Apple App Store or Google Play subscription settings.
13. Children
Agentale is not directed to children under 13 and does not seek a children’s-category listing. Users under 18 should review this Policy with a guardian, and anyone below the local age of independent digital consent may not create a cloud account independently. Child sexual exploitation and immediate child-safety risks receive priority blocking and escalation.
14. International transfers
Servers, AI models, or processors may be outside your country. Where required, we use lawful mechanisms such as assessments, certifications, standard contractual clauses, or adequacy decisions, together with minimization, encryption, and access control. Review processing regions and policies for providers you configure yourself.
15. Contact and updates
We may update this Policy as features, laws, processors, or practices change. Material changes receive prominent notice and renewed consent where required. The latest version and effective date appear above.
