1f:"$Sreact.view_transition" 20:I[9754,["/_next/static/chunks/2t36f5p4pjzlb.js","/_next/static/chunks/1gyqeidjm5eoa.js","/_next/static/chunks/0_fjnxc4tt_l_.js","/_next/static/chunks/1ymt1shyhmu-4.js","/_next/static/chunks/3kv2zf0i7s2wc.js","/_next/static/chunks/1vs_3g_g75zmk.js","/_next/static/chunks/2g4dh802s8oz3.js","/_next/static/chunks/1bc3f0shukf1j.js","/_next/static/chunks/192kgs9v_t14j.js","/_next/static/chunks/0z6pw0v5c_lx5.js","/_next/static/chunks/08r60e6icmtkc.js","/_next/static/chunks/3trx6c0i7_x-3.js","/_next/static/chunks/40sro5bzp9bi_.js","/_next/static/chunks/1wgf3tk38y6cf.js"],"ProjectDetail"] 21:T5325, vibenvr — 本地视频监控，隐私无忧

本地视频监控 · 隐私无忧

时间轴回放，
数据留在本地

vibenvr — 简单、隐私友好的本地 NVR 系统。快速搭建、灵活录制，无需云端。

统一事件时间轴按本地存储回放：每一帧都留在你自己的设备上，从未离开过你的网络。

24 小时事件带 — 时间刻度

统一事件时间轴高峰期 19:00–23:00 标注一次运动事件

/* 回放本地记录 — 每格一小时 */

采集 · 检测 · 持久化

PIPELINE

01 — 采集

接入 IP 摄像头

通过 ONVIF 与标准协议接入网络摄像头，构建统一事件时间轴。

02 — 检测

多引擎运动检测

同一画面可交给多种检测引擎，选择精度与算力的平衡。

03 — 持久化

写入本地存储

录制结果写入 SQLite 或 PostgreSQL，适合低功耗边缘设备。

三种检测引擎

ENGINES

引擎	作用	特点
OpenCV	基于像素差的标准运动检测，轻量、通用，适合大多数日常场景。	通用 · 算力友好
ONVIF Edge	在摄像头端做检测，仅在事件发生时上送，减少流量与功耗。	边缘 · 低带宽
AI · TFLite	TFLite 对象识别，区分“人 / 车 / 其他”，减少无关告警。	对象识别

存储账本

STORAGE LEDGER

事件 00

事件 01

事件 02

事件 03

当日保留 · 本地 00:11:35

录像与事件索引都保留在本地存储，不留存到外部云服务。

部署与访问控制

CONTROL

Docker 一键部署

容器化安装，快速上线；选择 PostgreSQL 或 SQLite 持久化。

docker compose up -d

安全认证

JWT 认证、2FA、RBAC 权限控制，谁看、看什么都可管理。

JWT · 2FA · RBAC

多语言界面

界面支持 10 种语言，包括中文、英文、日文等。

10 LANG

智能家居集成

通过 MQTT 与 Home Assistant Discovery 无缝对接。

MQTT · Home Assistant

本地视频监控，隐私无忧

vibenvr 把录制、检测与回放都留在你的网络里。数据不出本地，全部由你掌控。

MIT 许可 · JavaScript · 开源

22:T4ee1, vibenvr — local video surveillance, privacy-first

Local video surveillance · privacy-first

Your record plays back,
frame stays local

vibenvr — a simple, privacy-respecting local NVR. Fast setup, flexible recording, no cloud required.

The unified event timeline replays off local storage: every frame lives on your own device and never leaves your network.

24-hour event band — one windowed frame per hour

Unified event timelinePeak window 19:00–23:00 pins a motion event

/* replaying the local record — one hour per frame */

Capture · detect · persist

PIPELINE

01 — Capture

Bring in IP cameras

Attach network cameras over ONVIF and standard protocols into one unified event timeline.

02 — Detect

Multi-engine motion detection

Route the same feed to several detection engines and balance accuracy against compute.

03 — Persist

Write to local storage

Recordings are indexed in SQLite or PostgreSQL, suited to low-power edge devices.

Three detection engines

ENGINES

Engine	What it does	Profile
OpenCV	Standard pixel-difference motion tracking — lightweight and general, a fit for most everyday scenes.	General · compute-light
ONVIF Edge	Detection runs on the camera itself, with events forwarded only when they fire — less bandwidth, less power.	Edge · low-bandwidth
AI · TFLite	TFLite object recognition that tells people from vehicles and the rest, cutting irrelevant alerts.	Object recognition

Storage ledger

STORAGE LEDGER

Event 00

Event 01

Event 02

Event 03

Retained today · local 00:11:35

Recordings and event indexes stay on your local storage — nothing is parked with an outside cloud.

Deploy & access control

CONTROL

One-command Docker deploy

Containerized install to go live fast, with SQLite or PostgreSQL for persistence.

docker compose up -d

Hardened authentication

JWT auth, 2FA, and RBAC access control — who sees what stays under your hand.

JWT · 2FA · RBAC

Localized interface

The UI ships in 10 languages, including Chinese, English, and Japanese.

10 LANG

Smart-home integration

Talks to your home stack over MQTT and Home Assistant Discovery.

MQTT · Home Assistant

Local surveillance, privacy-first

vibenvr keeps recording, detection, and replay inside your network. The data never leaves — it stays fully in your hands.

MIT-licensed · JavaScript · open source

a:[["$","$1f",null,{"enter":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-enter","default":"none"},"exit":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-exit","default":"none"},"update":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-update","default":"none"},"default":"none","children":["$","$L20","cmt2zh30b01lx627108t5ska4:zh-CN:beautified",{"project":{"id":"cmt2zh30b01lx627108t5ska4","slug":"spupuz-vibenvr","name":"vibenvr","description":"简单、隐私友好的本地 NVR 系统","authorName":"spupuz","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":"https://vibenvr.org/","qualityScore":93,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zh30o01m96271h5exdt0p","kind":"artifact","artifact":{"schemaVersion":"localized-static-html-v4","kind":"static_html","locales":["zh-CN","en"],"defaultLocale":"zh-CN","documents":{"zh-CN":"$21","en":"$22"},"designDirections":{"schemaVersion":"dodot-design-directions-v1","directions":[{"id":"timeline-recall","name":"Timeline Recall","projectMetaphor":"A security timeline that plays back the local record — the strip of frames that proves what stayed on your storage, never sent to a cloud.","signatureVisual":"A ruled event timeline drawn in native CSS/SVG — a horizontal line with 24 tick marks, one windowed keyframe nodule per hour, and a highlighted interval where a motion event is pinned, all colored to suggest a frame playing under local control.","heroGeometry":"A wide horizontal timeline band across the hero, arithmetic of hours instead of marketing cards.","sectionSequence":["timeline","pipeline","detection-engines","storage-floor","control","privacy-line"],"surfaceModel":"Recorded-roll surfaces: layered paper frames with a subtle frame-ledge edge, each sheet a card that reads like an archived playback, separated by a thin ruled spine.","palette":"Deep ink-navy page, archival negative-tape amber and signal red accents, grey paper-bleed neutrals — the film-strip register.","typography":"Monospace tabular numerals for the time axis and load-bearing values, a condensed grotesk for headings, system neutral for body.","desktopComposition":"Six parts stacked in vertical reading order: the timeline hero, then an offset pipeline row, a three-purpose detection table, a storage ledger, a controls column, and a closing privacy rule.","mobileTransformation":"The wide hero timeline collapses into a 24-column mini strip; windowed nodules pinch into a dot row; ledger and controls stack into single tall columns.","mediaStrategy":"No media region — the timeline itself is the derived illustration of recording; pure typographic and data marks.","forbiddenPatterns":["hero slogan + three metric side-cards","feature-card wall that fits any SaaS","camera-viewfinder shell or dashboard chrome frame"],"compositionFingerprint":"wideline-hero-24hr-mini-strip-ledger-stack-nocolordrift"},{"id":"signal-island","name":"Signal Island","projectMetaphor":"An isolated island network where every camera signal stays local — a surveillance graph kept fully on the wire, none of it routed out to the cloud.","signatureVisual":"A hand-drawn concentric island diagram — lens nodes ringing a central storage hub, connected by curved local-sine links, one faint route dashed outward and barred with a slash to mark the 'no cloud' boundary.","heroGeometry":"Radial rings centered on a local storage core, not an empty viewfinder center.","sectionSequence":["lens-map","vertex-edges","engine-cores","volumes","guards","perimeter"],"surfaceModel":"Soft-ink map strata: vignette paper ground, tinted lens plaques on radial guide curves, guarded by a darker perimeter ring.","palette":"Warm map-atlas cream and sepia ground, lake-blue lens rings, one barred orange cloud route — a cartographic register.","typography":"Small-caps cartographic labels, ligature serif for headlines, mono for network coordinates and security parms.","desktopComposition":"Radial hero map on the left with a text column on the right, then perimeter-guarded sections radiating outward as the eye reads out from the core.","mobileTransformation":"The concentric map flattens into a linear chain — hub on the left, lenses stacked to the right, rings becoming bordered rows; the barred cloud route stays the closing motif.","mediaStrategy":"No remote media; the island diagram is the derived illustration of a sealed local network.","forbiddenPatterns":["grid-of-three number badges","faded camera-tile mosaic background","telemetry sparkline dashboard shell"],"compositionFingerprint":"radial-island-map-perimeter-guard-linear-chain-sigmaring"},{"id":"record-floor","name":"Record Floor","projectMetaphor":"A raw ledger of retained footage — the physical floor of storage where every gigabyte is accounted for and nothing is ever pushed offsite.","signatureVisual":"A vertical retention ledger built from native rows — each row a dated recording slice with an in-row motion spark and a filled bar showing its local volume share, all summed under a running total.","heroGeometry":"A stacked retention ledger column that doubles as the page spine, not a hero banner.","sectionSequence":["ledger","engines","rail","stores","locks","tally"],"surfaceModel":"Ribbed record-shelf surfaces: repeating ruled ledgers and ledger headers, register-marked columns, a stitched spine where sections meet.","palette":"Cold ink well and paper white, stamped black type, a single signal-red total, mineral gray fills — a printing-press register.","typography":"Tabular mono throughout the ledger, strong stencil-weight display for titles, light caption faces for metadata.","desktopComposition":"A single tall central ledger column with rotated marginal columns for engines, rails, stores, locks, and a pinned running total — the ledger is the spine the whole page hangs on.","mobileTransformation":"The ledger column narrows to fit one view; marginal columns fold into header strips stacked above each ledger block; the running total docks to the bottom edge.","mediaStrategy":"No media region; the retention ledger is the derived illustration of accountable local storage.","forbiddenPatterns":["three-stat hero topper","card wall divided by rounded rects","browser + terminal split mockup"],"compositionFingerprint":"ledger-spine-vertical-tally-docked-footer-stencilpress"}],"selectedDirectionId":"timeline-recall","selectionRationale":"All three directions are evidence-native to VibeNVR (motion detection, unified event timeline, local storage). timeline-recall is chosen because the project's core mechanism is the time-driven event record: cameras feed a unified timeline that stays on local storage. Its horizontal-time hero directly encodes 'unified event timeline', it has zero reliance on media or viewfinder chrome (no qualified local asset exists), it reads clean at both 1280 and 390 widths through a distinct 24-hour-strip collapse, and its sequence (timeline → detection engines → storage floor → controls → privacy) mirrors the real causal chain of capture, detect, persist, control, guard. It differs from signal-island (radial cartography) and record-floor (tall ledger spine) in metaphor, hero geometry, surface model, and mobile recomposition, so it is not a palette swap of the others."},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"timeline-recall","projectMetaphor":"A security timeline that plays back the local record — the strip of frames that proves what stayed on your storage, never sent to a cloud.","signatureVisual":"A ruled event timeline drawn in native CSS/SVG — a horizontal line with 24 tick marks, one windowed keyframe nodule per hour, and a highlighted interval where a motion event is pinned, all colored to suggest a frame playing under local control.","compositionFingerprint":"wideline-hero-24hr-mini-strip-ledger-stack-nocolordrift","heroGeometry":"A wide horizontal timeline band across the hero, arithmetic of hours instead of marketing cards.","sectionSequence":["timeline","pipeline","detection-engines","storage-floor","control","privacy-line"],"surfaceModel":"Recorded-roll surfaces: layered paper frames with a subtle frame-ledge edge, each sheet a card that reads like an archived playback, separated by a thin ruled spine.","palette":"Deep ink-navy page, archival negative-tape amber and signal red accents, grey paper-bleed neutrals — the film-strip register.","typography":"Monospace tabular numerals for the time axis and load-bearing values, a condensed grotesk for headings, system neutral for body.","desktopComposition":"Six parts stacked in vertical reading order: the timeline hero, then an offset pipeline row, a three-purpose detection table, a storage ledger, a controls column, and a closing privacy rule.","mobileTransformation":"The wide hero timeline collapses into a 24-column mini strip; windowed nodules pinch into a dot row; ledger and controls stack into single tall columns.","mediaStrategy":"No media region — the timeline itself is the derived illustration of recording; pure typographic and data marks.","evidencePlan":["Timeline hero encodes the verified unified event timeline via native hour ticks with a motion interval pinned","Detection table names the three verified motion engines (OpenCV, ONVIF Edge, AI/TFLite) as evidence rows","Controls and policy rows state only verified items: Docker deploy, JWT/2FA/RBAC, 10-language UI, local/edge storage"],"localeContract":{"zh-CN":"以本地回放时间轴隐喻组织章节；文案密度自然、使用中文排版；所有数值与引擎名仅使用已验证事实。","en":"Organized by the local-playback timeline metaphor; English copy keeps natural density; all figures and engine names stay within verified facts."},"forbiddenPatterns":["hero slogan + three metric side-cards","feature-card wall that fits any SaaS","camera-viewfinder shell or dashboard chrome frame","repeated identical radius/shadow/blur on every section","decorative gradient or particle field as the signature visual"]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"页面在 zh-CN 与 en 中一致地实现了所选 timeline-recall 方向：hero 是一条原生 HTML/CSS 绘制的 24 小时事件时间轴（每小时刻度、关键帧结节、19:00–23:00 高亮运动区间），六节顺序 timeline→pipeline→detection-engines→storage-floor→control→privacy-line 与 brief 完全一致。所有事实均有 material.json 依据（三种检测引擎、Docker/PostgreSQL/SQLite、JWT/2FA/RBAC、10 语言、MQTT/Home Assistant、MIT/JavaScript、本地存储无云）。390px 断点将时间轴坍缩为 24 列迷你条+圆点行，pipeline 纵向堆叠、检测表转为块状网格、账本与控件列收窄，构成真实重构。无通用 Hero/三卡/卡片墙/装饰渐变表单视觉，无假控件，无无依据设备壳。zh-CN 存在一处无害的失效 CSS 选择器（#why-xs 未在标记中使用），非阻断。","checks":{"projectFit":{"passed":true,"evidence":"hero 的宽横向 24 小时时间轴带正是 brief heroGeometry 'wide horizontal timeline band across the hero, arithmetic of hours'；HTML 六节 data-dodot-section 顺序 timeline→pipeline→detection-engines→storage-floor→control→privacy-line 与 design-brief.sectionSequence 逐一一致。"},"evidenceBound":{"passed":true,"evidence":"三种引擎 OpenCV/ONVIF Edge/AI TFLite 对应 material.summary evidence 第 2、3 条；Docker/PostgreSQL/SQLite 对应第 2 条；JWT/2FA/RBAC 对应第 4 条；10 语言含中英日对应第 6 条；MQTT/Home Assistant 对应第 5 条；MIT/JavaScript 来自 repository.license/language；'本地存储无云' 与 description 及 highlights 'Local storage, no cloud' 一致。账本时长(00:11:35)等为示意数据但仅在本地保留概念范围内。"},"signatureVisual":{"passed":true,"evidence":"signatureVisual 真实存在：24 个 hour-col 用 CSS grid repeat(24,1fr) 绘制，每小时 hour-rule+hour-lab+nodule，nodule.on/red 区分关键帧与运动区间，底部 .window 红色运动条 + tl-caption '19:00–23:00 标注一次运动事件'，全部为原生 CSS/DOM，非图片。"},"mobileTransformation":{"passed":true,"evidence":"@media (max-width:760px) 中 .timeline gap:2px 且隐藏 .hour-lab 收为圆点行、hour-col 定高 58px；.pipeline-row 由三列改 grid-template-columns:1fr 纵向堆叠；table.det 转为块状 1fr 1fr 网格；.ledger-row 收窄为 110px/1fr；.control-grid 改单列。属 brief 所述 'collapse into 24-column mini strip / stack single tall columns' 的真实重构。"},"localeParity":{"passed":true,"evidence":"zh-CN 与 en 均为 6 节同构，数据值一致（24 小时轴、19:00–23:00 运动窗、账本 30%/8%/55%/12% 与 00:11:35、三引擎、10 LANG、docker compose up -d、MQL 等）。翻译各自地道自然：中文用 '采集·检测·持久化''存储账本''本地视频监控，隐私无忧'，英文 'Capture·detect·persist''Storage ledger''Local surveillance, privacy-first'，长度匹配（en textLength 2451 vs zh 1123，符合语言密度差异）。"},"antiTemplate":{"passed":true,"evidence":"未回退通用模式：hero 无'slogan+三指标卡片'而是数据时间轴；中部用偏移 pipeline、三列检测表、存储账本、控件列，而非任意 SaaS 的卡片墙；无摄像头取景圆壳/仪表盘 chrome；signatureVisual 是数据驱动的 24 刻度时间轴而非装饰性渐变/粒子场。palette（ink-navy/amber/red）与 typography（mono 时间轴+grotesk 标题）落实 film-strip register。"},"noFakeUi":{"passed":true,"evidence":"无假控件：'docker compose up -d''JWT·2FA·RBAC''10 LANG''MQTT·Home Assistant' 均为静态文本栈，非可点击伪造控件；无假按钮/播放器；空链接为 0；goto 链接真实指向 github 与首页。时间轴 role=\"img\"+aria-label 是正确无障碍标记。"}},"issues":[{"code":"dead-css-wp-xs","severity":"warning","locale":"zh-CN","description":"zh-CN.html 第 61 行存在选择器 table.det td#why-xs{display:none}，但标记中没有任何 td 带有 id=\"why-xs\"，该规则不会命中，属失效 CSS（en 版省略了此规则）。不影响布局或内容。","repairInstruction":"删除该条失效规则，或若意图在移动端隐藏各引擎第二列则按实际类名（如 td.why）调整并确认不会误隐内容。"}]},"repairEvidence":[],"provenance":{"executor":"opencode","model":"deepseek/deepseek-v4-flash","opencodeVersion":"1.1.58","promptVersion":"dodot-showcase-deepseek-v4","inputHash":"df7ea4c27dc038263dc17a71979f4b5163bfdb7d91b007b1baea1ae7cd144049","artifactHash":"0762bc82fb26f0028f5698bab9f6d9973df4a465ca908ea5ae039b401ad42f2b","designSkill":"dodot-showcase-design"},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1123,"cjkCount":481,"latinLetterCount":227,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3912,"h1Top":199.984375,"bottomInset":274.921875,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"7662221ae62b17e5d54e16a1b5cc9f3c21e1635344689b39e87c6bb6c6443ad1","middle":"7637441c54867a540e4e4d9aee27022f717f11174b07378296afde3d8b080996","bottom":"16c29324498d55f5452d446e5ab38d4913cd160325402825857903c3736a6451"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1042,"cjkCount":475,"latinLetterCount":227,"scrollWidth":390,"clientWidth":390,"scrollHeight":4169,"h1Top":118.390625,"bottomInset":253.75,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"e74d7f06bc6deb73ade3fb5a967935ed12ebc2239111d409f9f6cabcab25783a","middle":"37fc642ddf45f169bc240bbde8748289a239328948bcf22ef82031b2fcaf5611","bottom":"69cf7db5af62c3fdacdda585a87f4c69a9d044ef7b5235817273a08ac181f471"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2451,"cjkCount":0,"latinLetterCount":1821,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4131,"h1Top":199.984375,"bottomInset":275.5625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"c228ed3e7779f8c02eb6175aa76057f16e502a922269fe8504c790a8d26cf441","middle":"1f8379fa4b1a7854f76bc0526bc2262cf2ff14c4d6bcd73b3d43b5622179a824","bottom":"351b5bc0d974c6395380c4fe1e5cc920b5d8418a32eddb0e5a024ed1b0783b59"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":2351,"cjkCount":0,"latinLetterCount":1798,"scrollWidth":390,"clientWidth":390,"scrollHeight":4983,"h1Top":138.78125,"bottomInset":253.578125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"3e0e1875400c191ce0b555f8fe6aaba477004fcdb3e6de7f4d6e40abbda5faf4","middle":"2e8c6dc72c5925a179fa93f1f94f592b0a9721f780c7f0de534a0493a7033b42","bottom":"d9ac0fd143f992e989c2de710e2bc96707eb677d44df4c3b09fff2e95a56476b"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T09:05:55.949Z"},"rawExtraction":{"projectRevisionId":"cmt2zh30h01m36271hgvxhv21","projectVersion":1,"sources":[{"canonicalKey":"github:spupuz/vibenvr","role":"primary","revisionVersion":1,"url":"https://github.com/spupuz/VibeNVR"}],"facts":[{"key":"repository_owner","value":"spupuz","evidence":[{"path":"repository.owner","excerpt":"spupuz"}]},{"key":"repository_name","value":"vibenvr","evidence":[{"path":"repository.name","excerpt":"vibenvr"}]},{"key":"description","value":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required","evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},{"key":"primary_language","value":"JavaScript","evidence":[{"path":"repository.language","excerpt":"JavaScript"}]},{"key":"license","value":"MIT","evidence":[{"path":"repository.license","excerpt":"MIT"}]},{"key":"topics","value":"ffmpeg, lightweight, local-storage, nvr, opensource, privacy, video-surveillance","evidence":[{"path":"repository.topics","excerpt":"ffmpeg, lightweight, local-storage, nvr, opensource, privacy, video-surveillance"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:spupuz/vibenvr"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-17T09:58:40.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},{"slug":"analyze_data","label":"分析与理解数据","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]},{"slug":"learn_explore","label":"学习与探索","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"Simple, privacy-respecting local NVR — fast setup, flexible recording, no cloud required"}]}],"readme":{"headings":["📹 VibeNVR – Modern Video Surveillance System","🎬 Video Demo","✨ Key Features","🔒 Security & Architecture","📊 Public Telemetry","What is collected?","How often?","Opt-out","🌍 Localization & Translations","🚀 Quick Start","Prerequisites","📦 Installation via Docker (Recommended)","🔄 Updating VibeNVR","💾 Data Persistence (Bind Mounts vs Volumes)","🏠 Dashboard Integration (Homepage)","🎥 Live View Streaming Architecture","🌐 Production Deployment (Nginx Proxy Manager)","🔧 Troubleshooting","Permission Errors on Proxmox/PVE Kernel, OpenMediaVault, Synology, QNAP","Enable unprivileged user namespaces"],"commands":["docker compose -f docker-compose.prod.yml up -d","docker compose -f docker-compose.prod.yml pull","docker compose -f docker-compose.prod.yml down","docker compose -f docker-compose.prod.yml up -d","docker compose down","docker compose up -d"]},"changeSet":[{"field":"repository","beforeHash":"791ae2d7cf194f23cbcd16fcf6938826e98017d531b975bba309d2f84d6c4f55","afterHash":"579af5fab4007cc6d3f34a0bc7e670d784a90bb22d87c8e4a3af6117d385211a"}]},"publishedAt":"2026-08-16T12:17:49.501Z","updatedAt":"2026-08-21T13:26:53.852Z","summary":{"id":"cmt2zh30k01m56271q0ukaiez","projectType":"product","oneLine":"简单、隐私友好的本地 NVR 系统","overview":"VibeNVR 是一个现代化的本地视频监控系统，支持 IP 摄像头管理、录制、运动检测和统一事件时间线。它提供 Docker 部署，支持 PostgreSQL 和 SQLite，适合低功耗边缘设备。内置多种运动检测引擎（OpenCV、ONVIF Edge、AI），并具备 JWT 认证、2FA、RBAC 等安全功能。界面支持 10 种语言，并集成 MQTT 和 Home Assistant。","highlights":["本地存储，无需云","多引擎运动检测","Docker 一键部署","支持 10 种语言"],"bestFor":["希望自建监控系统的用户","注重隐私的家庭用户","需要边缘部署的开发者"],"caveats":["需要 Docker 环境","AI 检测可能需要额外硬件","社区支持以英文为主"],"categorySuggestion":"productivity","subcategorySuggestion":"automation_workflows","lifecycleStages":["deployment_growth"],"audienceIntents":["evaluate","operate","build"],"confidence":0.8,"evidence":[{"claim":"VibeNVR 是一个本地 NVR 系统，支持 IP 摄像头管理、录制、运动检测和统一事件时间线。","sourceType":"readme","locator":"https://github.com/spupuz/VibeNVR"},{"claim":"提供 Docker 部署，支持 PostgreSQL 和 SQLite，适合低功耗边缘设备。","sourceType":"release","locator":"https://github.com/spupuz/VibeNVR/releases/tag/v1.33.0"},{"claim":"支持多种运动检测引擎：OpenCV、ONVIF Edge 和 AI（TFLite 对象识别）。","sourceType":"readme","locator":"https://github.com/spupuz/VibeNVR"},{"claim":"提供 JWT 认证、SSO/OAuth、2FA、RBAC 权限控制等安全功能。","sourceType":"readme","locator":"https://github.com/spupuz/VibeNVR"},{"claim":"支持 MQTT 和 Home Assistant Discovery，便于智能家居集成。","sourceType":"readme","locator":"https://github.com/spupuz/VibeNVR"},{"claim":"界面支持 10 种语言，包括中文、英文、日文等。","sourceType":"readme","locator":"https://github.com/spupuz/VibeNVR"}],"publishedAt":"2026-08-17T13:44:49.261Z"},"github":{"url":"https://github.com/spupuz/VibeNVR","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/1721792?v=4","starsCount":365,"forksCount":27,"language":"JavaScript","syncedAt":"2026-08-17T13:41:13.844Z","license":"MIT","defaultBranch":"main","archived":false,"disabled":false,"lastChangedAt":"2026-08-17T09:58:40.000Z","missing":false},"tags":[],"updates":[]},"entryPoint":"discovery","claim":{"state":"unclaimed","verified":false},"initialView":"beautified","community":{"viewer":null,"state":{"likesCount":0,"creativeCount":0,"followersCount":0,"commentsCount":0,"liked":false,"creative":false,"followed":false},"comments":{"comments":[],"nextCursor":null}}}]}],"$L23"]

时间轴回放，数据留在本地

采集 · 检测 · 持久化

接入 IP 摄像头

多引擎运动检测

写入本地存储

三种检测引擎

存储账本

部署与访问控制

Docker 一键部署

安全认证

多语言界面

智能家居集成

本地视频监控，隐私无忧

Your record plays back,frame stays local

Capture · detect · persist

Bring in IP cameras

Multi-engine motion detection

Write to local storage

Three detection engines

Storage ledger

Deploy & access control

One-command Docker deploy

Hardened authentication

Localized interface

Smart-home integration

Local surveillance, privacy-first

时间轴回放，
数据留在本地

Your record plays back,
frame stays local