Privacy Policy

Last updated: 2026-05-16

UniCone ("we", "us", or "our") develops and operates UniCone Agent (the "Product"). This Privacy Policy (the "Policy") explains how we collect, use, store, share, and protect your information when you use the Product or visit our official website.

Please read this Policy carefully before using the Product. By starting to use the Product, you acknowledge that you have read, understood, and agreed to the practices described in this Policy. If you do not agree with this Policy, please do not use the Product.

If you have any questions about this Policy, please contact us using the information at the end of this document.

Our Core Commitments

• We never sell your personal data, nor share it with third parties for advertising or marketing purposes.
• Your personal data will not be used to train AI models without your explicit consent.
• Screenshots and local sensing data are by default kept only on your local device and are not uploaded to our servers.
• You are always in control: you can pause or completely disable any context-capture capability (screen, clipboard, IM capture, voice, etc.) in the settings at any time.
• The Product adopts a "local-first" architecture by default: core data (todos, notes, calendar, conversation history, vector memory) can be stored entirely in a local database without being uploaded to the cloud.

Table of Contents

1. Scope
2. Information We Collect
3. How We Use This Information
4. AI Features and Third-Party Models
5. Local Sensing Capabilities and Screen / IM Context
6. Sharing and Disclosure of Information
7. Cookies and Similar Technologies
8. Data Storage Locations and Cross-Border Transfers
9. Data Retention
10. Data Security
11. Protection of Minors
12. Your Rights
13. Changes to This Policy
14. Supplementary Terms for Specific Jurisdictions
15. Contact Us

1. Scope

This Policy applies to:

• The UniCone Agent desktop application (based on Tauri / Electron for Windows / macOS);
• The UniCone official website;
• Mobile applications and hardware device integrations that work with the Product;
• Any local or private-cloud Server you choose to deploy.

This Policy does not apply to third-party services that require your separate authorization (for example, Alibaba Cloud Bailian, Google Gemini, ByteDance Doubao, Tavily, email service providers, etc.). When you integrate such third-party services with the Product, your relationship with that third party is additionally governed by its own privacy policy and terms of service.

2. Information We Collect

2.1 Information You Provide Voluntarily

• Account information: when you register or log in, you provide your email address, username, and password (stored encrypted).
• Profile information: nickname, avatar, time zone, language preference, and similar items that you voluntarily fill in under personal settings.
• Task and schedule content: todos, sub-tasks, notes, tags, due dates, calendar events, and other content you create.
• AI conversation content: messages exchanged with the Product's AI assistant, as well as user preferences and profile data ("Agno Learning" memory).
• Customer support information: content you actively provide when reporting issues, submitting tickets, or joining user communities.

2.2 Information Generated Through Your Use

• Task execution records: action logs such as checking off, modifying, moving, or archiving todos, used to provide task history and statistics.
• AI tool-call records: tool calls performed by the AI assistant to fulfill your requests, including confirmations (HITL) and results.
• Audio data (optional): voice snippets you actively record when voice input is enabled, together with their transcribed text.
• Local sensing data (optional, requires you to actively grant system permissions):
  • Screenshots and OCR text: used to build a "personal activity stream (LifeTrace)" and to provide work context to the AI.
  • Instant messaging capture (IM UIA): after you actively enable the "IM capture" toggle for a specific application (e.g., WeChat, Feishu, DingTalk), the Product can read the currently visible message text from that chat window for scenarios such as "extract todos from chat".
  • The above sensing data is by default kept only in the database on your local device (SQLite + ChromaDB) and will not be uploaded to our servers without your active authorization.
• Integration and connector data (optional): relevant metadata and necessary content from mailboxes (IMAP/SMTP), third-party task systems, file directories, etc., that you actively configure and authorize.

2.3 Automatically Collected Information

• Device information: operating system type and version, application version, device model, CPU/GPU architecture, screen resolution, time zone, and language.
• Usage data: aggregated statistics such as feature usage frequency, session duration, and interaction patterns (when anonymized analytics is enabled).
• Error and diagnostic logs: when the Product crashes, encounters an exception, or experiences a call failure, error stack traces, call stacks, error codes, and related request summaries are generated automatically. We make reasonable efforts to anonymize or redact personal information in these logs.
• Observability data: model-call latency, token consumption, and billing data (when you have enabled the Cloud API model catalog), used for billing, rate-limiting, and capacity planning.

2.4 Information We Do Not Actively Collect

• We do not actively read the content on your screen unless you have explicitly granted system-level screen recording / accessibility permissions and have enabled the corresponding context-capture switch in the Product settings.
• We do not read your keyboard keystrokes, nor do we record your mouse movements.
• We do not access your contacts, SMS, camera, or other system capabilities unrelated to the Product's features without your permission.

3. How We Use This Information

We use the information we collect for the following purposes:

• Providing core features: to present and sync your todos, schedules, AI conversations, and personal profile.
• AI assistance: to enable intelligent task breakdown, context-aware suggestions, the conversational AI assistant, memory, and cross-session learning.
• Local sensing: to build a "personal activity stream" within the scope of your authorization, providing work context to the AI assistant so it can more accurately understand your current task.
• Account and security: for registration, sign-in, identity verification, fraud prevention, and abuse prevention.
• Operations and improvement: to diagnose issues and improve product performance and experience through aggregated, anonymized usage data.
• Customer support: to respond to your tickets and answer your inquiries.
• Legal compliance: to process necessary data to fulfill legal obligations when required by applicable laws or competent authorities.

Without your separate, explicit consent, we will not use your personal data to:

• Train our own AI models;
• Train third-party AI models;
• Make automated decisions that produce legal or similarly significant effects on you.

4. AI Features and Third-Party Models

The AI capabilities of the Product rely on large language models (LLMs) and related multimodal models. Model providers and processing arrangements are as follows:

4.1 Default Routing Through the "UniCone Private Cloud" Gateway

By default, the Product calls LLMs through an OpenAI-compatible gateway provided by UniCone. In this mode, your conversation requests are first sent to UniCone's private-cloud relay service, which then forwards them to the specific model provider.

4.2 Third-Party Model Providers You May Configure Yourself

The Product supports configuring your local API key for any OpenAI-compatible provider. When you actively configure and use such a provider, your relevant data (including conversation text, authorized file content, authorized screen OCR text, etc.) will be transmitted to that provider for processing. Currently known configurable providers include, but are not limited to:

| Provider | Default Use | |---|---| | Alibaba Cloud Bailian (DashScope, Qwen + DeepSeek series) | Primary chat model, small models, vision models | | Google Gemini | Diary illustration generation, optional multimodal capabilities | | ByteDance Doubao (Volcano Ark / Doubao Seedream) | Optional image generation | | OpenAI-compatible providers such as OpenAI / Anthropic / DeepSeek | Configured manually by you | | Alibaba Cloud ASR (DashScope WebSocket) | Optional real-time speech recognition | | Tavily Search | Optional web search |

Please review each third-party service's privacy policy before configuring it. We forward requests to a provider only when you have actively configured and use the corresponding feature; otherwise, we do not actively send your data to any of the providers listed above.

4.3 Data Sent to Model Providers

When you trigger an AI feature, only the minimum context necessary for that request is sent to the model provider, which may include:

• The prompt and conversation messages you currently input;
• Todos, notes, or file snippets you actively select and attach;
• Text that has been captured with your active authorization and processed via OCR (raw screenshots themselves are not sent to cloud model providers, unless you actively choose to upload images in a scenario that explicitly supports multimodal input);
• Necessary system prompts and tool definitions.

We use technical and contractual measures to make reasonable efforts to ensure that model providers do not use your data for training. However, please be aware: the actual data processing by third-party providers on their platforms is governed by their own privacy policies and terms of service.

4.4 You May Disable AI Features at Any Time

• You can leave the LLM API key blank in the settings, in which case all AI-related features become unavailable, while local features such as todos and calendar continue to work normally.
• You can disable background tasks that trigger model calls, such as "automatically extract todos from screen / IM".
• You can toggle individual capabilities: screenshots, IM capture, voice input, email integration, web search, and other optional features.

5. Local Sensing Capabilities and Screen / IM Context

The desktop application of the Product includes built-in local sensing capabilities — optional modules that run on your local device to perform screenshots, OCR, IM capture, and similar tasks, providing "work context" to the AI assistant so that it can understand what you are currently doing. This section describes our principles for handling such sensitive data.

5.1 Data Collection Requires Your Dual Authorization

Local sensing capabilities collect data only when both of the following conditions are met:

1. You have actively granted the corresponding system permissions in your operating system ("Screen Recording / Accessibility" on macOS, the corresponding permissions on Windows);
2. You have actively enabled the corresponding capture toggle in the Product settings (screenshots, IM capture, clipboard, voice, etc., can be toggled independently).

If either condition is not met, the corresponding sensing capability will not run.

5.2 Data Is Kept Locally by Default

• Screenshots: by default stored in the local application data directory and not uploaded to our servers.
• OCR text: generated locally by an OCR engine on your device (such as RapidOCR / WinRT OCR) and stored in a local SQLite database.
• IM capture text: only within the scope of applications you have actively enabled (e.g., if only WeChat is enabled and DingTalk is not, only WeChat is read), the text of the currently visible chat is read via system-level UI Automation interfaces and stored in the local database.
• Vector memory (ChromaDB): embedding vectors derived from the above local data are also stored locally.

5.3 The Only Path Through Which Data Reaches the Cloud

Only when you actively trigger an AI feature and that feature genuinely requires relevant context will the Product select the minimum necessary text snippets (not the raw data) from local data and send them, together with your current request, to the model provider described in Section 4. We do not silently upload local sensing data to our servers in the background.

5.4 Recommendations for Sensitive Scenarios

To maximize the protection of your privacy, we strongly recommend that you:

• Temporarily disable the screenshot and IM capture toggles when handling highly sensitive content such as password input fields, bank account details, medical records, or legal documents;
• Regularly review and clean up the local sensing database;
• Enable sensing features only on trusted devices and networks.

5.5 Deleting Local Data

You can delete local sensing data through any of the following means:

• Click the "Re-run initialization wizard" button in the settings to clear memory data, contacts profiles, IM capture data, and speaker profiles in one go;
• Delete the Product's local data directory directly;
• When uninstalling the Product, follow the standard uninstall procedure on the installation platform to clean up associated data.

6. Sharing and Disclosure of Information

We do not sell your personal data, nor do we use it for cross-context behavioral advertising. We share information only in the following limited circumstances:

• AI model providers: as described in Section 4, only when you actively use AI features, the minimum necessary request content is transmitted to the model provider you have configured.
• Infrastructure service providers: for example, cloud databases, cloud storage, content delivery, logging and crash analytics, email delivery, etc. We bind these providers via data processing agreements that allow them to process data only on our instructions.
• Legal requirements: we may disclose necessary information in accordance with applicable laws, regulations, court orders, or lawful requests from competent authorities, or to protect the legitimate rights and interests of the Product, its users, or the public.
• Business changes: if the Product undergoes a merger, acquisition, reorganization, or asset sale, the relevant user data may be transferred as part of the transaction; we will notify you in an appropriate manner before such transfer takes effect.
• Anonymized and aggregated data: we may conduct internal research and product improvement based on data that has been anonymized and aggregated, which does not constitute a "sale" of personal information.

7. Cookies and Similar Technologies

On our official website, we use cookies and similar technologies to:

• Maintain login status and sessions;
• Remember your theme, language, and other preferences;
• Understand site usage via anonymized analytics to improve product experience.

You can manage or disable cookies through your browser settings; disabling certain cookies may affect the availability of related features. The desktop application does not rely on web cookies to maintain login state.

8. Data Storage Locations and Cross-Border Transfers

• Local storage: core business data (todos, notes, calendar, conversation history, sensing data, vector memory) is stored by default in a database on your local device.
• Self-hosted Server: if you choose to deploy the Server on your own servers or private cloud, the relevant data is kept on infrastructure you control.
• UniCone hosted service: if you use the hosted service provided by UniCone, the relevant data may be stored on cloud-provider infrastructure within mainland China.
• Third-party model providers: when you use AI features, the relevant requests may be transmitted to the country or region in which the model provider is located (for example, the United States, Singapore, the EU, or mainland China) for processing.

Where cross-border data transfers are involved, we will adopt protective measures consistent with this Policy. If your jurisdiction has specific requirements for cross-border data transfers (such as the standard contractual clauses under the GDPR or the cross-border compliance requirements of the PIPL), please assess the deployment regions of the model providers involved before use.

9. Data Retention

We retain your personal data only for the shortest period necessary to achieve the purposes described in this Policy, unless a longer retention period is required or permitted by law:

• Account information: retained throughout the validity of your account; deleted within 90 days after account closure.
• Task and conversation content: retention period is consistent with your account; you may manually delete individual records within the Product at any time.
• Local sensing data (screenshots / OCR / IM): kept only on your local device; you may clean it up manually at any time, and the Product also provides an optional toggle for periodic automatic cleanup.
• Error and diagnostic logs: typically no more than 90 days, used to diagnose and fix issues.
• Aggregated / anonymized usage data: may be retained long-term in anonymized form for product analytics.
• Raw audio data: deleted upon completion of local transcription, unless you actively choose to retain the voice snippet.

10. Data Security

We adopt commonly used industry technical and organizational measures to protect your personal data, including but not limited to:

• Transport encryption: communications with our servers use TLS (HTTPS / WSS).
• Storage encryption: sensitive fields (such as authentication credentials) are stored in salted-hashed or encrypted form in the database.
• Principle of least privilege: internal systems use role-based access control and multi-factor authentication.
• Change auditing: writes to critical data and configuration changes leave audit logs.
• Dependency governance: dependencies are regularly scanned and upgraded for vulnerabilities.

Please note that no internet transmission or electronic storage can be 100% secure. If you discover or suspect a security vulnerability in the Product, please contact us immediately using the information at the end of this Policy.

11. Protection of Minors

The Product is not directed to minors under 18 years of age. We do not knowingly collect personal information from minors. If your jurisdiction sets a higher minimum age (for example, 16 years in some EU member states), the higher age shall prevail.

If you are a parent or guardian and become aware that a minor has submitted personal information without your consent, please contact us, and we will delete the relevant information promptly in accordance with the law.

12. Your Rights

To the extent permitted by applicable law, you have the following rights with respect to your personal information:

• Right to be informed and right of access: to know what data we process about you and to request access to it.
• Right to rectification: to request correction of inaccurate or incomplete information.
• Right to erasure: to request deletion of your personal information when conditions are met (including account closure).
• Right to data portability: to obtain a copy of your data in a structured, commonly used, machine-readable format.
• Right to restriction / right to object: to object to or restrict our processing of your data in specific circumstances.
• Right to withdraw consent: to withdraw consent you previously gave at any time; withdrawal does not affect the lawfulness of processing based on consent before the withdrawal.
• Right to lodge a complaint: to lodge a complaint with the data protection authority in your jurisdiction.

Please exercise the above rights through the contact information at the end of this Policy. We will respond to your request within the period required by applicable law (typically no later than 30 days).

13. Changes to This Policy

We may update this Policy from time to time. When we make material changes, we will notify you through in-app notifications, email, or announcements on the official website. The "last updated" date at the top of this Policy always reflects the latest version. Your continued use of the Product after the changes take effect constitutes your acceptance of the revised Policy.

14. Supplementary Terms for Specific Jurisdictions

14.1 Users in the People's Republic of China

If you are located within the People's Republic of China, the processing activities described in this Policy are subject to laws and regulations such as the Personal Information Protection Law, the Data Security Law, and the Cybersecurity Law. You have the right to request that we disclose our personal information processing rules and provide explanations, and you may file complaints with competent authorities such as the Cyberspace Administration, MIIT, or public security organs.

14.2 Users in the EU / EEA / UK / Switzerland

Within the scope of the GDPR and the UK GDPR, you enjoy all data subject rights under the GDPR and have the right to lodge a complaint with the data protection authority in your member state (for example, EU: ec.europa.eu/justice/data-protection/bodies/authorities; UK: ico.org.uk; Switzerland: edoeb.admin.ch).

14.3 Users in California, USA, and Other Similar Jurisdictions

Within the scope of laws such as the CCPA / CPRA, you have the right to know what information we collect and share, to request access to, correction of, and deletion of your personal information, and to opt out of "sale" or "sharing". We do not sell your personal information, nor do we use it for cross-context behavioral advertising.

15. Contact Us

If you have any questions, comments, or requests regarding this Policy or our data processing, please contact us through the following channels:

• Email: public@freeyou.club
• Phone: +86 176 0627 5971
• Official website: unicone.club

We take every piece of feedback seriously and will respond within a reasonable time.