Spark KNE Verify Privacy Policy

Last updated: August 10, 2026

Spark KNE Verify provides user-initiated knowledge assessment through browser extensions and approved platform integrations. This page is static HTML so platform reviewers and users can read the full policy without running JavaScript. It explicitly describes data collection, processing and use, storage and retention, sharing and disclosure, user deletion, contact, and security.

Chrome Web Store Review Summary

Collection: Spark collects account email, account ID, authentication state, anonymous trial identifier, usage quota, selected verification text, source page URL, page title when available, source-platform label, client channel, contribution content, feedback, blocking issue reports, consented interaction events, aggregate homepage install-link clicks, timestamps, trust-state results, review status, quota counters, diagnostic events, IP address, user-agent, and operational logs.
Processing and use: Spark uses this data for login, beta access, anonymous trial verification, authenticated API calls, selected-text verification, source/evidence comparison, trust-state output, consented interaction-funnel measurement, aggregate install attribution, optional Verify history, contribution review, feedback review, issue debugging, quota enforcement, reliability, security, fraud prevention, and abuse prevention.
Storage: Chrome extension storage is used for local settings, language, anonymous trial identifier, authentication token, account state, usage display, latest verification state, consent state, interface preferences, and a temporary retry queue for consented interaction events that failed to reach Spark. Spark servers store account, anonymous trial usage, Verify history, contribution, feedback, issue report, quota, review, analytics, source gap, and operational records.
Sharing: Spark does not sell, rent, trade, or share user data with advertisers or data brokers. Data is sent only to Spark-controlled services at https://knespark.com and infrastructure providers needed to host, secure, monitor, back up, and operate the service.
Deletion: users can remove local extension data by logging out, removing the extension, or clearing Chrome extension storage. Server-side deletion requests can be sent to seth@knespark.com.

Prominent Disclosure and Consent

Spark sends data for verification only after you take an explicit action, such as selecting text and clicking Verify, using the context menu, copying a feedback package, reporting a blocking issue, or submitting a contribution.
Version 0.1.8 shows a data-use disclosure, privacy-policy link, and consent checkbox inside the extension before anonymous trial verification, interaction analytics, login, registration, authenticated verification, feedback, issue reporting, debug-token saving, or KNE contribution can send data.
By using anonymous trial Verify, signing in, using Verify, Submit to KNE, feedback, or issue reporting, you consent to the collection, use, storage, and sharing described in this policy.

OpenAI and Coze Platform Integrations

When you explicitly invoke Spark KNE Verify inside an approved OpenAI or Coze experience, that platform sends the claim text and optional source URL you provided to Spark over HTTPS. Spark returns the trust state, explanation, evidence counts, and public source references to the same platform.
The current OpenAI integration is anonymous and read-only. It does not request an OpenAI account identifier, Spark account, Verify history, contribution access, or KNE write permission.
The current Coze integration uses a server-held platform key that is not exposed to end users.
Spark applies Redis-backed platform quotas and may process IP address, User-Agent, request time, endpoint, status, source platform, client channel, optional source URL, and a limited claim excerpt in operational verification records for quota enforcement, security, debugging, service measurement, and abuse prevention.
OpenAI and Coze independently process the conversation and returned tool result under their own privacy terms. Spark does not send platform-integration data to advertisers or data brokers.

Data We Collect

Account data: email address, account ID, plan type, public beta status, quota state, login state, and role.
Anonymous trial data: a locally generated anonymous trial identifier, hashed server-side identifier, anonymous trial quota state, and anonymous verification usage. Anonymous trial data is not counted as a registered public beta user.
Authentication data: access tokens stored in Chrome extension storage and sent to Spark over HTTPS for authenticated API requests.
Website content and personal communications: only the text you actively select or submit for verification, including AI chat text or other conversation text if you choose to verify it, plus source page URL, page title when available, and source platform label such as ChatGPT, Qwen, Claude, Doubao, Gemini, or web. Technical request metadata also identifies the client channel, such as Chrome extension, Edge extension, OpenAI MCP, or Coze plugin.
User activity: Verify requests, contribution submissions, feedback submissions, blocking issue reports, interaction events, timestamps, quota counters, trust-state results, source hint counts, review status, and diagnostic events needed to operate the beta.
Consented interaction analytics: after consent, Spark records extension open, selection captured, Verify clicked, verification result viewed, and first Verify outcome. Interaction metadata contains the entry point, selection presence, a coarse text-length bucket, request identifier, trust state, and verification context as applicable; it does not include selected text or a page URL.
Homepage install attribution: clicking an install link on knespark.com records only the site locale, link placement, and campaign tags. This event does not include selected text or an account identifier.
User-provided content: source URLs, source hints, corrections, challenges, notes, feedback text, issue report text, and selected text that you choose to submit.
Location data: Spark does not collect GPS, precise location, or address data. Server and security infrastructure may process IP address as operational log data for security, reliability, fraud prevention, abuse prevention, and troubleshooting.
Technical and operational data: client channel, extension or integration version, manifest version, build ID, package timestamp, request time, endpoint, status code, error category, IP address, user agent, security logs, and abuse-prevention signals.

How We Use Data

To provide anonymous trial verification before login, limited to a small quota window.
To authenticate your account and keep you signed in.
To verify selected text against KNE sources, evidence candidates, trust policy, and review rules.
To return trust states, source details, evidence gaps, and risk warnings.
To store optional Verify history and help you revisit past verification results.
To operate beta quotas, detect abuse, debug errors, monitor reliability, and improve the product.
To measure consented interaction events and recover their delivery after temporary network failures.
To review user-submitted sources, corrections, challenges, feedback, and issue reports.
To measure aggregate public beta behavior such as activation, repeat Verify rate, source gaps, and blocker trends.
To measure aggregate homepage install attribution across entry points and campaigns.

How We Store and Retain Data

Local storage: the extension stores API settings, UI language, anonymous trial identifier, authentication token, account state, usage display, latest verification state, consent state, interface preferences, and a temporary interaction event retry queue in Chrome extension storage.
Retry queue: if a consented interaction event cannot reach Spark after a temporary network failure, the extension may retain up to 30 event records locally for retry. These records contain event type, source-platform label, client channel, extension version, event identifier, and interaction metadata; they do not contain selected verification text or a page URL. Delivered records are removed; older records are removed when the queue limit is reached.
Server storage: Spark servers may store account records, anonymous trial usage records, Verify history, contribution records, feedback, issue reports, quota counters, review status, analytics events, source gap tasks, and operational logs.
Retention: account, Verify history, contribution, feedback, and issue report records are retained while your beta account is active or while needed to provide the service, review contributions, improve KNE quality, ensure security, resolve disputes, prevent abuse, or meet legal obligations.
Operational logs are retained only as long as needed for reliability, debugging, security, abuse prevention, or legal obligations.
You can clear local extension data by logging out, removing the extension, or clearing Chrome extension storage.

Sharing and Disclosure

Spark does not sell, rent, or trade user data.
Spark does not share selected text, account data, authentication data, Verify history, feedback, or contributions with advertisers or data brokers.
Data is transmitted to Spark-controlled services at https://knespark.com.
For an approved OpenAI or Coze integration, Spark receives the user-requested claim from that platform and returns the resulting trust state and public evidence references to the same platform. Spark does not forward that claim or result to unrelated platforms.
Current infrastructure processors include Vultr cloud servers and the browser/Chrome extension storage environment controlled by the user. These processors may handle data only as needed for hosting, database operation, HTTPS delivery, logging, security, monitoring, backup, and service reliability.
Reviewed contribution content may be used inside KNE source and evidence workflows, but it does not directly modify KNE Core data without review.
Spark may disclose data if required by law, security investigation, abuse prevention, dispute resolution, or to protect the service and users.

Limited Use

Spark's use and transfer of information received from Google APIs will adhere to the Chrome Web Store User Data Policy, including the Limited Use requirements. Spark does not use Chrome extension user data for advertising, sale, resale, creditworthiness, lending, or unrelated third-party profiling.

Security

Data sent between the extension and Spark servers uses HTTPS.
Authentication tokens are stored locally in Chrome extension storage and are not sold or shared with advertisers.
Server-side access to beta data is limited to service operation, debugging, abuse prevention, security, and KNE review workflows.

What Spark Does Not Do

Spark does not monitor your clipboard.
Spark does not record browser history in the background.
Spark does not continuously read AI conversations.
Spark does not collect passwords, payment card numbers, government identifiers, health records, or precise location.
Spark does not sell personal data.

User Choices and Deletion

You control when text is sent by choosing Verify, Submit to KNE, feedback, or issue reporting.
You can stop collection by logging out, not using Verify or Submit to KNE, disabling the extension, or removing it from Chrome.
To request account deletion, Verify history deletion, contribution deletion, tester feedback deletion, issue report deletion, or privacy support, contact seth@knespark.com.
Deletion requests are handled by deleting or anonymizing relevant account and service records unless retention is required for security, abuse prevention, dispute resolution, or legal obligations.

Contact

For privacy or support questions, contact seth@knespark.com.

Spark KNE Verify 隐私政策

最后更新：2026 年 8 月 10 日

Spark KNE Verify 通过浏览器插件和获准的平台集成，为用户主动提交的知识主张提供可信状态评估。本页面是静态 HTML，平台审核人员和用户无需运行 JavaScript 即可读取完整隐私政策。本页面明确说明数据收集、处理和使用、存储和保留、共享和披露、用户删除、联系方式和安全措施。

Chrome Web Store 审核摘要

收集：Spark 会收集账号邮箱、账号 ID、认证状态、匿名试用标识、用量额度、选中的验证文本、来源页面 URL、可用时的页面标题、来源平台标记、客户端入口、贡献内容、反馈、阻断问题上报、已同意的交互事件、官网安装链接汇总点击、时间戳、可信状态结果、复核状态、额度计数、诊断事件、IP 地址、User-Agent 和运维日志。
处理和使用：Spark 将这些数据用于登录、公测访问、匿名试用验证、已认证 API 请求、选中文本验证、来源/证据比对、可信状态输出、已同意的交互漏斗统计、官网安装归因、可选 Verify 历史、贡献复核、反馈复核、问题排查、额度执行、稳定性、安全、防欺诈和防滥用。
存储：Chrome 扩展存储用于保存本地设置、语言、匿名试用标识、认证 token、账号状态、用量显示、最近验证状态、同意状态、界面偏好，以及未能发送至 Spark 的已同意交互事件临时重试队列。Spark 服务器保存账号、匿名试用用量、Verify 历史、贡献、反馈、问题上报、额度、复核、分析、来源缺口和运维记录。
共享：Spark 不出售、出租、交易用户数据，也不会把用户数据分享给广告商或数据经纪商。数据只会发送到 Spark 控制的 https://knespark.com 服务，以及托管、安全、监控、备份和运行服务所必需的基础设施提供方。
删除：用户可以通过退出登录、移除插件或清理 Chrome 扩展存储删除本地扩展数据。服务端删除请求可发送至 seth@knespark.com。

显著披露与同意

Spark 只会在你明确操作后发送数据，例如选中文本并点击 Verify、使用右键菜单、复制反馈包、上报阻断问题或提交贡献。
0.1.8 版本会在插件内匿名试用验证、交互统计、登录、注册、已登录验证、反馈、问题上报、保存调试 token 或 KNE 贡献发送数据前，显示数据使用说明、隐私政策链接和同意复选框。
你使用匿名试用 Verify、登录并使用 Verify、Submit to KNE、反馈或问题上报，即表示你同意本政策说明的数据收集、使用、存储和共享方式。

OpenAI 与 Coze 平台集成

当你在获准的 OpenAI 或 Coze 体验中明确调用 Spark KNE Verify 时，该平台会通过 HTTPS 把你提供的主张文本和可选来源 URL 发送给 Spark。Spark 会把可信状态、说明、证据计数和公开来源引用返回同一平台。
当前 OpenAI 集成为匿名只读工具，不会请求 OpenAI 账号标识、Spark 账号、Verify 历史、贡献访问权或 KNE 写入权限。
当前 Coze 集成使用保存在服务端的平台密钥，密钥不会暴露给最终用户。
Spark 使用 Redis 持久化平台额度，并可能在运维验证记录中处理 IP 地址、User-Agent、请求时间、接口、状态码、来源平台、客户端入口、可选来源 URL 和有限长度的主张摘录，用于额度执行、安全、排错、服务衡量和防滥用。
OpenAI 和 Coze 会依据各自的隐私条款独立处理对话和返回的工具结果。Spark 不会把平台集成数据发送给广告商或数据经纪商。

我们收集的数据

账号数据：邮箱、账号 ID、计划类型、公测状态、额度状态、登录状态和角色。
匿名试用数据：本地生成的匿名试用标识、服务端 hash 标识、匿名试用额度状态和匿名验证使用记录。匿名试用不计入注册公测用户。
认证数据：保存在 Chrome 扩展存储中的访问 token，并通过 HTTPS 发送给 Spark 用于已登录 API 请求。
网页内容和个人通讯：只有你主动选择或提交用于验证的文本，包括你选择验证的 AI 聊天文本或其它对话文本，以及来源页面 URL、可用时的页面标题、ChatGPT、千问、Claude、豆包、Gemini 或普通网页等来源平台标记。技术请求元数据还会标记 Chrome 扩展、Edge 扩展、OpenAI MCP 或 Coze 插件等客户端入口。
用户活动：Verify 请求、贡献提交、反馈提交、阻断问题上报、交互事件、时间戳、额度计数、可信状态结果、来源线索数量、复核状态和运行公测所需诊断事件。
已同意的交互统计：同意后，Spark 会记录插件打开、捕获选区、点击 Verify、查看验证结果和首次 Verify 结果。交互元数据按事件需要包含入口、是否存在选区、粗粒度文本长度区间、请求标识、可信状态和验证语境，不包含选中文本或页面 URL。
官网安装归因：点击 knespark.com 的安装链接时，只记录网站语言、入口位置和活动标签；该事件不包含选中文本或账号标识。
用户提供内容：你主动提交的来源 URL、来源线索、纠正、质疑、备注、反馈文本、问题上报文本和选中文本。
位置信息：Spark 不收集 GPS、精确位置或地址数据。服务器和安全基础设施可能把 IP 地址作为运维日志处理，用于安全、稳定性、防欺诈、防滥用和排错。
技术与运维数据：客户端入口、插件或集成版本、manifest 版本、build ID、打包时间、请求时间、接口、状态码、错误类型、IP 地址、User-Agent、安全日志和防滥用信号。

我们如何使用数据

用于账号认证和保持登录状态。
用于将选中文本与 KNE 来源、证据候选、可信策略和复核规则进行验证。
用于返回可信状态、来源详情、证据缺口和风险提示。
用于保存可选 Verify 历史，帮助你回看历史验证结果。
用于执行公测额度、防止滥用、排查错误、监控稳定性和改进产品。
用于统计已同意的交互事件，并在临时网络故障后恢复事件发送。
用于复核用户提交的来源、纠正、质疑、反馈和问题上报。
用于汇总分析公测行为，例如激活率、二次验证率、来源缺口和阻断问题趋势。
用于汇总比较官网不同安装入口和活动的获客归因。

我们如何存储和保留数据

本地存储：插件会在 Chrome 扩展存储中保存 API 设置、界面语言、匿名试用标识、认证 token、账号状态、用量显示、最近一次验证状态、同意状态、界面偏好和交互事件临时重试队列。
重试队列：已同意的交互事件因临时网络故障未能发送至 Spark 时，插件最多会在本地保留 30 条事件记录用于重试。这些记录只包含事件类型、来源平台标记、客户端入口、插件版本、事件标识和交互元数据，不包含选中的验证文本或页面 URL。成功发送后会删除对应记录；达到队列上限时会移除较早记录。
服务器存储：Spark 服务器可能保存账号记录、Verify 历史、贡献记录、反馈、问题上报、额度计数、复核状态、分析事件、来源缺口任务和运维日志。
保留期限：账号、Verify 历史、贡献、反馈和问题上报记录会在公测账号有效期间，或在提供服务、复核贡献、改进 KNE 质量、保障安全、处理争议、防止滥用或满足法律义务所需期间保留。
运维日志只在稳定性、排错、安全、防滥用或法律义务所需期间保留。
你可以通过退出登录、移除插件或清理 Chrome 扩展存储删除本地扩展数据。

共享与披露

Spark 不出售、出租或交易用户数据。
Spark 不会把选中文本、账号数据、认证数据、Verify 历史、反馈或贡献分享给广告商或数据经纪商。
数据会传输到 Spark 控制的服务： https://knespark.com 。
对于获准的 OpenAI 或 Coze 集成，Spark 从该平台接收用户主动请求验证的主张，并把可信状态和公开证据引用返回同一平台；不会把该主张或结果转发给无关平台。
当前基础设施处理方包括 Vultr 云服务器和用户控制的浏览器/Chrome 扩展存储环境。这些处理方只会在托管、数据库运行、HTTPS 传输、日志、安全、监控、备份和服务稳定性所需范围内处理数据。
通过复核的贡献内容可能用于 KNE 来源和证据工作流，但不会未经复核直接修改 KNE Core 数据。
如法律要求、安全调查、防止滥用、争议处理或保护服务与用户所需，Spark 可能披露相关数据。

Limited Use

Spark 对从 Google API 获得的信息的使用和转移将遵守 Chrome Web Store 用户数据政策，包括 Limited Use 要求。Spark 不会将 Chrome 扩展用户数据用于广告、出售、转售、信用评估、借贷或无关第三方画像。

安全

插件和 Spark 服务器之间的数据传输使用 HTTPS。
认证 token 保存在 Chrome 扩展存储中，不会出售或分享给广告商。
服务端公测数据访问仅限服务运行、排错、防滥用、安全和 KNE 复核流程。

Spark 不会做的事

不监听剪贴板。
不在后台记录浏览历史。
不持续读取 AI 对话。
不收集密码、银行卡号、政府身份号码、健康记录或精确位置。
不出售个人数据。

用户选择与删除

你通过选择 Verify、Submit to KNE、反馈或问题上报来控制何时发送文本。
你可以通过退出登录、不使用 Verify 或 Submit to KNE、停用插件或从 Chrome 移除插件来停止数据收集。
如需删除账号、删除 Verify 历史、删除贡献内容、删除测试反馈、删除问题上报或获得隐私支持，请联系 seth@knespark.com。
删除请求会删除或匿名化相关账号和服务记录，除非安全、防滥用、争议处理或法律义务要求继续保留。

联系方式

隐私或支持问题请联系 seth@knespark.com。