Stop handing your daily food diary, body weight history, and biometric sleep data to ad-funded fitness clouds: NutriTrace brings complete calorie, macro, micronutrient, and wellness tracking to your own server with an offline-capable web PWA and native Android app. Log meals across customizable slots using live Open Food Facts and USDA FoodData Central lookups, barcode scanning, AI nutrition label OCR, or direct recipe imports from Mealie and CookTrace. Calorie and macro targets adjust automatically through three goal engines, including an inspectable Adaptive TDEE algorithm that regresses 35 days of intake against scale weight at 7,700 kcal per kilogram. Connect Fitbit via Google Health, Garmin, Withings, and Android Health Connect to pull steps, heart rate variability, sleep stages, and body composition into unified Sleep, Readiness, and Resilience scores. An integrated Trace AI assistant and Smart Log let you log meals by voice, text, or meal photo, while the opt-in Model Context Protocol server exposes 20 read, write, and management tools to Claude Desktop, Cursor, and Codex. Multi-user households get OpenID Connect SSO and one-step migration from MyFitnessPal, Cronometer, Lose It, and Waistline. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

When an AI agent breaks in production, Tracely freezes that exact run into a replayable regression test and blocks the pull request that would ship the bug again. Span soup hides the failure. OTLP/HTTP traces group into conversation threads, and a waterfall marks the failing span in red with its input and output. Nobody knows if an answer was good. Evaluators sit as columns on the trace table, mixing model-free checks with LLM judges at conversation, run, or span level. Thirty-one broken runs, one issue. Clustering turns them into a single counted issue and can suggest the evaluator that would have caught it. A fix nobody can prove. One click promotes a failing trace into a hermetic case with recorded fixtures and a fail-to-pass contract, and scenarios send a red-team model after an adversarial goal. Model bills on every CI run. The suite replays offline with no API keys and no model spend, and tracely gate posts a commit status, comments on the pull request, and blocks regressions. No warning when it breaks again. Alert flows drawn on a canvas react to failed gates, broken conversations, unseen failure modes, or rising failure rates by messaging Slack, sending email, calling a webhook, or prompting an LLM. Coding agents can read traces and write evaluators through an MCP endpoint. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Mapping every physical server, virtual machine, container, and network link across a sprawling homelab into a live interactive canvas, Circuit Breaker gives infrastructure operators real-time topology visualization, IP address management, and rack elevation modeling on a single screen. Operators can switch between hierarchical, concentric, cortex ring, radial, and force-directed graph layouts where animated edges reflect link speeds and color-coded status rings display live hardware health. Automated discovery scans local subnets using nmap, SNMP, and ARP while querying Proxmox VE clusters, TrueNAS storage pools, and UniFi controllers, staging newly detected nodes in a review queue before adding them to the map. For multi-site labs or isolated VLANs, lightweight remote agents connect outbound over Noise-encrypted links after fingerprint approval to report host telemetry and run scoped network scans without opening inbound firewall ports. Rack diagrams let administrators drag servers into U slots with cabling overlays using a built-in catalog covering Dell, HPE, Supermicro, Ubiquiti, MikroTik, Synology, and APC gear alongside live iDRAC, iLO, and UPS metrics. Role-based access control, OIDC single sign-on, TOTP multi-factor authentication, encrypted full-state snapshots, and a SHA-256 hash-chained audit log keep multi-user operations accountable. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Instead of wrestling with raw command-line flags and YAML configuration files for every device on your self-hosted Headscale mesh VPN, Headplane gives you a Tailscale-style web administration console to manage machines, users, DNS records, and access policies from any browser. From the machines overview, network administrators can inspect connected nodes, view Tailnet IPv4 and IPv6 addresses, rename devices, reassign ownership across users, toggle key expiration, and approve advertised subnet routes or exit nodes in a couple of clicks. Built-in Access Control List (ACL) management lets you edit security policies, define user groups, assign node tags, and configure auto-approvers without manually SSHing into your control server. When someone needs immediate shell access to a remote Linux host on the mesh, the integrated WebSSH console opens an interactive terminal session right inside the browser tab over Tailscale SSH. You can also manage MagicDNS settings, configure split DNS nameservers, and add custom DNS records that automatically reload your Headscale instance. Multi-user administration supports OpenID Connect (OIDC) single sign-on with identity providers like Okta, Authentik, Authelia, and Google Workspace, alongside granular role assignments and pre-authentication key generation for unattended server enrollment. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Behind a single dashboard, MyIP brings together multi-source IPv4 and IPv6 geolocation lookups, WebRTC and DNS leak detection, global latency probes, and Autonomous System routing analysis so you can audit network privacy and connectivity without bouncing between a dozen ad-ridden diagnostic sites. Side-by-side IP cards compare your detected address across multiple independent providers, displaying country, city, local time zone, ASN details, RPKI prefix announcements, and upstream transit paths to Tier 1 backbones. Privacy tests immediately expose real IP addresses leaked over WebRTC SDP handshakes or misconfigured VPN DNS resolvers, while a browser fingerprint analyzer and a 258-item cybersecurity checklist help harden client configurations. Network engineers can test reachability and multi-round minimum latency across up to 60 websites using curated category packs, verify split-tunnel proxy rules, and run worldwide ping, MTR traceroute, and censorship checks powered by distributed Globalping nodes. Additional lookup tools include a multi-country DNS resolver for spotting DNS poisoning, Whois and MAC vendor queries, a local IPv4 and IPv6 subnet calculator, live status feeds for major cloud services, and shareable diagnostic reports exported as expiring links, Markdown, or JSON. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

One OpenAI-compatible endpoint, 1,600+ models behind it: Portkey Gateway sits between your app and every major LLM provider, deciding in real time where each request should go. When a provider times out or returns errors, automatic retries with exponential backoff kick in, and fallback rules quietly move traffic to your second-choice model before users notice anything. Load balancing spreads calls across multiple API keys and providers by weight, conditional routing sends each request down a different path based on its content or metadata, and per-request timeouts cut off providers that hang. Guardrails inspect inputs and outputs as traffic passes through: block banned words, screen for sensitive content, and enforce the checks your compliance team asked for, with dozens of pre-built rules ready to attach. Text is only the start: the same API shape reaches vision models, speech-to-text, text-to-speech, image generation, and realtime websocket APIs, so multimodal features stop multiplying your integration work. Routing policies live in JSON configs you attach per request or per client, which means experiments, cost controls, and provider migrations ship as configuration changes, not redeploys. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Commercial flight logbooks frequently lock historical travel records behind annual subscriptions or shut down unexpectedly, whereas AirTrail preserves your complete aviation history on an interactive world map under your own control. Enter a flight number to look up routes automatically or log departures manually with seat numbers, cabin classes, aircraft registrations, timestamps, notes, and custom metadata fields. Switching from another tracker takes seconds: the built-in importer ingests CSV, ICS, and JSON exports from MyFlightradar24, Flighty, App in the Air, OpenFlights, TripIt, JetLog, byAir, and JetLovers. The vector map renders geodesic flight paths, recorded GPS tracks, and airport runway details across customizable basemaps. Detailed statistics dashboards break down total distance flown, time in the air, top airlines, aircraft models, and visited countries. Families and travel groups can share a single instance with role-based permissions, OAuth single sign-on, and read-only public links, while developers automate entries through the versioned REST API or Model Context Protocol server. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. GPL-3.0 licensed.

Picture your object storage scattered across a home server, a friend's rack, and a cheap VPS in another country, still answering S3 requests like nothing happened: that is exactly the deployment Garage was built for. Every object is replicated to three zones, so losing a node, a disk, or an entire site degrades the cluster instead of destroying your data. Applications talk to it with the standard S3 API: create buckets, upload objects in multiple parts, list and copy them, and hand out access keys that are scoped per bucket. A separate web endpoint serves static websites straight from buckets with CORS support, which covers personal sites, documentation, and release downloads without another server in the chain. The whole system ships as one dependency-free binary that asks for about a gigabyte of RAM and any x86-64 or ARM CPU from the last decade, second-hand hardware very much included. There is deliberately no consensus protocol to babysit: nodes gossip their cluster layout, tolerate 200-millisecond links, and keep serving reads while parts of the fleet are unreachable. Deuxfleurs runs it in production for their own hosting, and common pairings include restic or rclone backup targets, Matrix media storage, and Nix binary caches. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Whether operating from a home shack, a portable Parks on the Air activation, or a multi-operator club station, Wavelog gives amateur radio operators a browser-accessible logbook for recording and confirming QSO contacts from anywhere. Real-time transceiver frequency and mode data flow directly into the Live QSO logging interface from Hamlib, FLRig, and WSJT-X through the companion WavelogGate helper or REST API endpoints. An integrated DX Cluster and visual band waterfall display active spots filtered by continent, band, and DXCC status so operators can click a spot and QSY their rig immediately. Built-in two-way synchronization handles QSL confirmations and log uploads across Logbook of The World (LoTW), eQSL, QRZ.com, HRDLog, ClubLog, and the QO-100 DX Club alongside ADIF 3.1.8 and Cabrillo file imports and exports. Interactive Maidenhead grid maps, satellite pass tracking, and automated progress matrices for over 20 international awards (including DXCC, WAS, WAC, WAZ, IOTA, SOTA, POTA, and WWFF) show exactly which bands and modes still need confirmation. Dedicated contest logging windows and role-based club station permissions let teams work the same event simultaneously without log conflicts. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Instead of running an Elasticsearch or Loki cluster, VictoriaLogs stores and searches your logs from a single zero-config server. Ship from Filebeat, Fluent Bit, Fluentd, Logstash, Vector, Promtail, Telegraf, the OpenTelemetry Collector, or the Datadog Agent. Syslog, journald, and Splunk endpoints work too. Query with LogsQL: word, phrase, and prefix filters combine with pipes that extract fields from raw lines and compute stats, top values, and facets. See results as a table, grouped by stream, as raw JSON, or as a live tail. Chart them in Grafana, or work from a terminal with vlogscli and pipe output into grep, less, or jq. Events with hundreds of fields are supported. So are high-cardinality fields like trace_id and user_id. Late, out-of-order backfills are accepted. During an incident, stream_context pulls the lines around any match. Tenants are keyed by AccountID and ProjectID headers. vmalert evaluates LogsQL queries as alerting and recording rules. CEF messages sent over Syslog are parsed automatically for SIEM work. Retention defaults to 7 days and can be capped by disk space. Per-day partitions snapshot and copy out with rsync or rclone. Cluster mode splits ingestion, querying, and storage across nodes for logs that outgrow one server. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

BrightBean Studio provides an open-source social media management suite that centralizes multi-account publishing, visual calendar scheduling, idea tracking, and cross-channel analytics into a unified self-hosted dashboard. Content creators and marketing agencies draft cross-network posts within an interactive composer that provides real-time channel previews and custom caption overrides for individual networks. The visual drag-and-drop calendar maps publishing queues across recurring weekly time slots, while an integrated Kanban board organizes brainstorming concepts from unassigned ideation to production-ready schedules. Community managers monitor unified social inboxes aggregating direct messages, mentions, and comments across Facebook, Instagram, LinkedIn, YouTube, TikTok, Bluesky, Threads, Pinterest, and Mastodon. Team leads configure multi-tier approval workflows with threaded editorial comments and issue passwordless 30-day magic links so external clients can approve drafts without registration. Built-in performance dashboards aggregate engagement rates, follower growth curves, and watch time metrics directly from native network APIs without relying on third-party data aggregators. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Preventing silent link rot across research libraries, LinkAce captures, indexes, and permanently archives web discoveries into searchable multi-user collections with automated Wayback Machine snapshots. Curators bookmark articles, documentation, tools, and media using custom browser bookmarklets or quick-add input fields that automatically scrape page titles and metadata descriptions. The built-in link monitor continuously checks saved endpoints for HTTP status anomalies, alerting curators whenever destinations move, throw server errors, or vanish from the live internet. For long-term preservation, the system dispatches automated backup requests to the Internet Archive Wayback Machine to capture historical copies of critical research material. Curators organize their library using nested lists and granular tag hierarchies, switching seamlessly between compact tables, detailed reading feeds, and visual card layouts. Teams and households can create separate user accounts to collaborate on shared collections, delegating read-only or editing privileges per list. The platform also generates individualized RSS feeds for public or private bookmark streams, synchronizes external data through a tokenized REST API, and dispatches automated database dumps to AWS S3 storage buckets. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. GPL-3.0 licensed.

No external database cluster to babysit, no proprietary cloud lock-in, and no bloated workspace hierarchy: Jotty stores every note, checklist, and Kanban board as plain Markdown and JSON files inside a single folder on your server. Need a quick grocery list with nested categories, progress bars, and drag-and-drop reordering? Spin one up in seconds, or flip any list into a multi-column Kanban project board with custom statuses, due-date calendar views, and a built-in time tracker. Write technical documentation in a TipTap-powered WYSIWYG and Markdown editor featuring syntax-highlighted code blocks, GitHub-flavored callouts, embedded Draw.io and Excalidraw diagrams, and OpenPGP client-side encryption for sensitive records. Connect ideas across your workspace using Obsidian-compatible wikilinks, inline mention suggestions, backlinks, and an interactive 2D and 3D knowledge graph that maps relationships between tasks and notes. Share individual notes or live Kanban boards publicly via customizable read-only links, or start the optional Model Context Protocol (MCP) companion container so AI assistants like Claude and Cursor can query your notes and check off items through your scoped REST API key. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

If your books live in a plain-text Beancount ledger, Fava turns that file into a browsable dashboard of statements, charts, and a built-in editor. Income Statement, Balance Sheet, Trial Balance, Journal, Holdings, and Statistics reports render as tables and charts, with stacked or single bars, monthly breakdowns, and at-cost conversion. Three top-bar filters narrow every report. Match by date, account pattern, #tag, ^link, payee, metadata, or amount; spaces require all conditions, commas accept any. The Query page runs Beancount Query Language statements like bean-query. Two-column results become line or treemap charts, and every result exports to CSV, plus XLSX and ODS with the Excel extra. The browser editor autocompletes accounts, payees, and tags. Press n for a quick-add form that records a new transaction. Drag a receipt onto an account name to file it in that account's documents folder, or onto a Journal row to attach it to the transaction. Budget directives from daily to yearly show remaining amounts in charts and reports. Colored indicators flag passed or failed balance checks on opted-in accounts. Importers feed a review list; extensions add custom report pages. Gmail-style keyboard shortcuts speed up daily use. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

To uncover deep viewing insights and AI-powered content recommendations across your self-hosted Jellyfin media server, Streamystats delivers a comprehensive analytics and watch-history platform. Once connected to one or more Jellyfin instances using native MediaBrowser authentication, the dashboard tracks live playback sessions, per-user watch durations, device and client breakdowns, and library-wide growth metrics without requiring third-party reporting plugins. You can import existing historical logs from Jellystat or the Playback Reporting plugin so years of viewing habits carry over immediately, then explore interactive watch-time graphs, genre radar charts, actor and director leaderboards, and a yearly Wrapped summary. When you enable vector embeddings through OpenAI, OpenRouter, Google Gemini, Voyage AI, or a local Ollama model, the recommendation engine calculates cosine similarity against your watch history and explains exactly which watched movies or series inspired each suggestion. An integrated AI chat assistant equipped with thirteen specialized function-calling tools lets you search your catalog semantically, ask natural-language questions about your viewing statistics, and generate shared movie recommendations for multiple household members at once. Administrators can also mark titles as watched or unwatched with automatic watch-time inference, manage watchlists over an external API, and connect Janitorr for automated media cleanup. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Turn your scattered collection of liquor bottles and bar tools into a master mixology station with an intelligent digital cellar. Bar Assistant tracks your personal liquor cabinet, inventory levels, and available mixers against a catalog of more than 500 cocktail recipes. The platform calculates exact alcohol by volume ratings, computes ingredient costs per serving, and dynamically highlights every cocktail you can craft with on-hand supplies. Built-in recipe scrapers extract formula details, measurements, and glassware requirements directly from web links, JSON payloads, or YAML files. Multi-bar workspaces allow home mixologists and commercial venues to separate private collections from communal drink menus while handling complex ingredient substitutions, custom infusions, and shopping lists. Integrated Meilisearch indexing powers instant typo-tolerant searches across spirit categories, flavor tags, and cocktail strengths. Visual cocktail cards display high-resolution drink photography, user ratings, glassware requirements, and step-by-step garnish instructions. Deploy Bar Assistant on RepoCloud with dedicated VPS compute, full root SSH credentials, and in-browser console access, backed by an open-source MIT license.

Organizations seeking complete data sovereignty over their documents, spreadsheets, and shared project folders can deploy OpenCloud as a self-hosted workspace for file sync, sharing, and real-time collaboration across web, desktop, iOS, and Android clients. Teams organize departmental assets inside dedicated Spaces where files belong to the project rather than individual user accounts, ensuring continuity when personnel change and enabling delegated space managers to assign granular read, edit, or upload permissions. Through built-in WOPI integration with Collabora Online, multiple colleagues can co-edit Microsoft Office and OpenDocument files simultaneously in the browser with automatic conflict locking and version history recovery. External partners and clients can access files via password-protected public links with custom expiration dates or upload deliverables into write-only File Drop folders without seeing existing directory contents. Cross-organization collaboration works seamlessly over the Open Cloud Mesh federation protocol, while full-text and tag-based search locates documents across personal and shared drives in seconds. Client devices synchronize changes in real time with support for virtual files on demand, resumable chunked uploads, and offline mobile access with camera document scanning. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Every batch a craft business ships carries a recipe version, a cost, and a trail of ingredient lots, and Craftplan, an open-source ERP, keeps all three connected. Receiving stock against a purchase order records each supplier lot code and optional expiry date, and partial deliveries keep the order open. Customer orders appear on a delivery calendar and can generate invoices. Scheduled items wait on the production board as unbatched cards until Batch All creates one batch with a frozen copy of the recipe. Capacity warnings flag products over their daily limit, and a forecast highlights materials that will run short. Finishing a batch consumes ingredient lots earliest-expiry first, skips any lot on hold or rejected, and snapshots material, labor, and overhead costs. Recipes are versioned, so past costs never change. A blend made in-house can be stocked as its own material, with its actual yield, lot costs, and source lots intact. Allergens and nutrition facts roll up from each material through every recipe, and completed batches print A5 or compact labels with a scannable barcode. The Trace Center follows a supplier lot forward to every customer who received it, or a finished batch back to its sources. Recall mode can place the lot on hold, and the trace exports as CSV or prints for the recall file. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.