Your company pays for AI subscriptions nobody can see, scattered across personal accounts and team plans, and Subpool is the self-hosted control plane that pools that authorized capacity and hands every employee one key with limits you govern. Administrators combine subscription and API accounts into pools behind a single managed endpoint, then allocate employee-specific keys across whichever provider accounts are healthy. The console watches account health, remaining subscription capacity, and when each plan window resets, and routing prefers subscription capacity first, falling back to paid API accounts under the very same employee key when a window runs dry. Rate-limit, expire, or revoke any person's access independently, and keep every key-to-account assignment visible and auditable instead of buried in a shared credential nobody can trace. The employee API speaks OpenAI-compatible Responses and Chat Completions, so existing tools point at your pool unchanged, while aggregate input and output token telemetry is tracked per key in provider-native units. For OpenAI accounts, a Costs API estimate shows the remaining grant balance; Bailian Token Plan accounts display their five-hour, weekly, or monthly quota windows. Upstream credentials stay encrypted, and prompts, responses, and source code are never persisted. One Go service with an embedded console and PostgreSQL runs from Docker Compose, with sign-in on port 8081 and the employee API on 8080. 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.

Your pi agent is mid-refactor on the server at home, you are on a train with only a phone, and Pi-Remote is the self-hosted relay that puts that live session in your browser so the work does not wait for your desk. Output, tool calls, and status changes sync in real time, with incremental catch up after a reconnect, and the session itself never leaves the local pi process. Send messages while pi is busy and they queue or steer, with delivery confirmed; switch the model or thinking level, compact the context, fork or clone a session, and walk the session tree, all executed locally. Several machines share one relay, and the UI lists each device, its online state, and its sessions, and can start a new session on the device you pick. Link a device by scanning a QR code, checking the 2 digit code on the confirmation page, and approving with the admin token. Desktop and mobile get separate layouts with six skins in Cursor, Claude, and Codex looks, dark and light, and token and cost usage updates in the status area. At home the relay can run serverless over the LAN, with a Cloudflare Tunnel switched on when you leave; on a server it runs under systemd or Docker Compose, and a device stays reachable only while its attach window is open. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Google Browser Scraper is a self-hosted SERP API that gets Google results from a real Chrome browser and your own residential proxy, because plain HTTP scraping is refused and a fresh browser on a fresh IP is refused almost as fast. It warms a sticky proxy session with ordinary browsing first, lets the browser pass Google's checks, and then switches to a hybrid mode that fetches results pages directly over HTTPS with the browser's own cookies, at about 0.4 MB and 1.4 seconds per search on a warm IP. Every page comes back as SERP-style JSON with organic results, ads, related questions, related searches, pagination, the AI Overview, and local results. If a proxy IP starts serving poisoned pages for some other query, the scraper catches it within a few pages, drops that IP, and asks the query again on a fresh one. Captchas and blocks are reported as captchas and blocks, never disguised as an empty result set, so your pipeline knows when to retry or rotate. Run it as a CLI for keyword files, import it as a Python library, serve it as a local HTTP API, or expose the MCP server so AI agents get live Google search. Every run ends with a traffic and cost summary per thousand pages, and any sticky residential proxy works, with NodeMaven sessions handled for you. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

ComfyUI's node graph is where workflows are built, not where pictures get made, and HEISS UI is the calm studio that sits on top of your ComfyUI server for the daily prompt, wait, look, tweak loop. Thirty model families work out of the box, from SD 1.5 and SDXL to Flux.2, Qwen-Image, Krea 2, and Wan 2.2, and missing text encoders, VAEs, or node packs each get a Download button, with Get all for the lot. Pick 2K and the model draws at 2048 in one run with no upscale in between, or send any finished picture to 2K or 4K with SeedVR2 in one click. Import your own workflow and it becomes clean controls over the same graph, with usual inputs connected and guesses marked for review. Paint over part of a picture and say what it becomes, and only that part changes, with inpainting across 18 image families and no extra nodes. The gallery auto-groups ten takes of one idea into a single tile, and opens on your phone while your rendering machine does the work. A Hidden vault locks private images behind Touch ID, Windows Hello, or a password; new PNGs can carry Civitai-ready metadata, and Zen mode shows only prompt and picture. Run it beside ComfyUI with Docker or npm on port 8787, or point it at ComfyUI on another machine. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Flopwire lets your AI coding agents message each other and search each other's history, so a Claude Code session on your laptop can ask the Codex session on your server what it already tried before duplicating the work. Agents see who is online and what they started on with peers, then reach each other mid-turn with send; a message arrives inside the recipient's running turn, or with its human's next prompt, and only your own live session can be woken deliberately. A device agent indexes transcripts from Claude Code, Codex, Devin CLI, and opencode within about a second; grep runs regex over every session, and search returns ranked matches naming the user, device, harness, session, and byte range each hit came from. From any full commit ID, trace finds the session that produced it and lets an agent read the conversation behind the change. The same capabilities ship as MCP tools and a messaging skill, so agents recover earlier decisions without you relaying context by hand. Secrets are redacted on the device before upload and again on the server, and path rules keep chosen repositories out of the index. One machine needs no server at all; add the team server with Docker Compose, Postgres, and any S3 compatible object store when a whole team's devices should share one searchable archive over TLS. 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.

Recurring chores should get cheaper every time you run them, and Umpteenth is a self-hosted agent service built on that idea: describe a job once in plain language, and each run teaches it to need less help. A job starts on a schedule, a webhook call, the API, or the Run now button, and each run gets a fresh disposable container destroyed afterwards. The first run is an Explore run where the agent works out the steps; afterwards a reflection step reads the transcript and writes short learnings plus reusable toolkit scripts into a numbered playbook version. Later runs are Assisted, starting from those learnings, and after three successful assisted runs that take the same steps, the job graduates: reflection writes a main script, and scripted runs execute it alone, spending no model tokens unless a check asks for a judgment call. If a scripted check fails, the agent takes over in the same sandbox, finishes the job, and repairs the script. The job page charts cost per run dropping toward zero. Connect HTTP MCP servers such as GitHub or Slack; model keys and MCP credentials stay on the host and never enter the sandbox. Each job sets its own schedule, model, reach, and per run spending limit. Deploy with Docker Compose, or on Kubernetes with the Helm chart, where every sandbox is a pod. 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.

One endpoint for every model account you own: 9Router Pro sits between your applications and a fleet of LLM providers, decides which account answers each request, and keeps the whole operation visible on a live dashboard. Its signature trick is prompt injection scoped three ways, globally, per API key, or per model, delivered either as a system prompt or as a scripted chat history that primes the model far harder than instructions alone. Keys are products here: each carries a rolling token budget over 24 hours, 7 days, or 30 days plus rate limits per minute, hour, and day, so a resold plan enforces itself with polite 429 responses. Routing stays healthy on its own, with accounts tiered as clean, suspect, or cooldown, least-recently-used spreading across the fleet, flaky upstream responses retried against the next account, and proxy pools health-checked with one-click cleanup of the dead. The monitoring dashboard leads with VPS specs, fastest and most-used providers, top API key, and usage spectrum charts, while backup and restore cover providers, combos, keys, and settings section by section. Everything runs as a single application with one SQLite file and a dashboard localized into 34 languages. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Search is being answered by AI crawlers first and humans second, yet most site owners cannot name a single bot that read their pages last week; Geo-Monitor answers that from the access logs you already have. A small collector sits beside Nginx or OpenResty, mounts the log read-only, and streams only new lines in batches, resuming after outages and surviving log rotation without ever touching your site's code. The console separates today, the last week, and the last month into identified requests, crawled pages, active crawlers, and trend lines, then drills into every visit with time, path, method, status code, claimed identity, and verification state. Identification and verification are deliberately different things here: a user agent only proves what a request claims to be, while a verified badge requires trusted log addresses and official IP evidence to agree. Twenty built-in identities cover OpenAI, Anthropic, Perplexity, Kimi, and Bytespider crawlers, and you can add, edit, or disable rules as new bots appear. Your sitemap syncs or imports as the content catalog, so coverage views show which pages crawlers actually reached, and collector health exposes heartbeat, backlog, and parsing problems. 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 homelab eventually earns the question "so what is actually running on all those boxes," and Homelab Atlas turns the answer into a diagram worth posting. Add your servers, the services grouped on each one, and the network gear between them, and the canvas lays everything out automatically with real application logos instead of gray rectangles. Searchable logo pickers work offline, uploads accept your own PNG, JPG, WebP, or SVG marks, and Arrange mode lets you drag and resize cards when the automatic composition is not the story you want to tell. Discovery spares you the typing: point it at Docker or import a saved scan, then review every found device and app, all starting unchecked, before anything joins the diagram as one undoable change. Connections express logical relationships and VLANs, hardware details ride on each card, and your project autosaves in the browser with a downloadable JSON file as the durable copy. When it is time to share, export previews render PNG or SVG with addresses, URLs, and emails hidden by default, plus literal terms you name yourself. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Close the laptop, walk away, and your coding agents keep working; CodeRail is the remote control that lets you supervise them from a browser or your phone without handing your machines to anyone's cloud. A small bridge on each computer connects outbound to your server, so there are no open ports and no VPN to babysit, and each machine keeps its own Claude Code and Codex logins because CodeRail never holds your AI keys. Chats show plans, tool calls, models, modes, and background tasks, and when an agent wants to run something sensitive, the approval lands with you wherever you are, including as a phone push notification. Real terminals outlive the browser tab, files can be viewed, edited, uploaded, and downloaded in place, and git work, pull requests included, sits beside the conversation that produced it. Each chat even gets a browser on the computer itself, shared live with everyone in the chat and drivable by the agent. Roles run Owner, Operator, or Viewer per computer, narrowed down to folders and agents, with every change audited, and sessions keep running with every browser closed while history stays readable offline. 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.

Enterprises keep buying AI tools and wondering why output does not compound, when the real gap is governance: nobody can say which agent owns what, what it may touch, or what it cost. Ybkkasi answers those three questions in one console for every MCP server, skill, agent, and AI application you run. Register each asset and move it through a lifecycle state machine from discovery to retirement, with 85 permission points deciding who can view, call, publish, or approve at every step. Work itself is organized as goals that break into projects and tasks, assigned to digital employees who each carry a home organization, a human supervisor, and a roster status. Approvals route through a dedicated center where lifecycle changes and sensitive actions wait for a human decision instead of happening silently. Every call is metered, so usage and cost reports attribute spend by department, by agent, and by skill, making showback a query rather than an argument. The audit log records who did what to which asset and when, giving compliance a trail they can actually read. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Your application already knows how to send mail through Amazon SES; NitroMTA lets it keep that code, change the endpoint and keys, and deliver from IP addresses you actually control. The engine answers a submission only after the message is committed to disk, so accepted mail survives a crash, a power cut, and even the loss of its host when the disk survives. Capture mode runs the whole engine locally and shows each message in a browser at port 9090 instead of sending it, which makes testing painless before any DNS exists. Out in production, new IPs warm up receiver by receiver from fifty messages a day, rising after clean days and backing off when complaints, bounces, or throttling climb. Each provider gets its own connection limits and cooldowns, hard bounces and spam complaints suppress the address automatically, and outgoing mail is DKIM-signed and sent over TLS with MTA-STS and DANE verified where published. Templates, identities, configuration sets, tenants, dedicated IPs, and SES-shaped events through an SNS subset cover the SES surface your tooling expects, and a CLI explains why any given message is still waiting. 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.

Most AI tools give one person one chat window; Nautilo is built for the room where your whole team, and each person's own AI Genie, work on the same thing at the same time. Give your Genie a personality, memory, face, and voice, then bring colleagues and their Genies into a shared Room to pull an idea apart, draft it, and send an agent off to research the missing piece. Six bundled creative apps cover the work itself: Writer for documents, Sheets for spreadsheets, Slides for decks, Board for planning, Design for visuals, and Video for clips, all editable by people and agents side by side. When the agent's draft misses, you grab the controls and rewrite the paragraph or move the lettering yourself instead of begging a prompt for a three-inch nudge. You choose the models and plug in your own provider API keys, connect tools through MCP, and let Genies operate connected workstations when a task needs real computer use. Membership, access, and tool permissions stay under your server's rules, and web, desktop, and mobile clients all connect to the same self-hosted instance. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Your inference server should not sit mostly idle while a model crawls along at ten tokens per second, so ReadyLLM gives you a visual way to deploy, watch, and tune LLMs until the hardware is actually earning its keep. Add the machine under your desk or a remote box over SSH, let hardware detection read its GPU, VRAM, CPU, and memory, and get model recommendations that fit what you own instead of wishful downloads. A model store catalogs text, image, and video models with parameter counts, precision, resolutions, and real weight filenames, background downloads from multiple sources, and integrity checks when they land. One click starts or stops llama.cpp, vLLM, or SGLang services, installs a missing engine matched to your GPU backend, and leaves runtime parameters open for manual tweaks. Live curves track GPU utilization, memory, throughput, and latency as requests flow. When speed disappoints, two-phase auto tuning searches the knobs that matter, checks VRAM feasibility before it wastes your time, and benchmarks honestly against latency, throughput, or prefill goals; an AI tuning mode can reason through the same loop across several rounds. Beyond chat models, ComfyUI generation covers text-to-image and text-to-video, storyboards, shot-by-shot long video, and upscaling, with prompts optionally expanded first. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Git servers still think in folders on one machine's disk, and Enroute is the Git server rebuilt like a modern database: your repositories live as immutable layers in object storage, served by compute nodes you can add or replace at will. Point any Git client at it and push; the push creates the repository, gets indexed and validated without locks, and commits once its entry is replicated across the ledger nodes' disks, which is how a single repository absorbs thousands of pushes per second. Clones and fetches always read one consistent snapshot, so readers never block writers, and background layer merging replaces the git repack jobs you used to schedule. The programmable part is hooks: your own code decides who can read or write, which branches a client may see, whether a ref update is allowed, and what happens after a push, from running CI to mirroring elsewhere. A rich gRPC API lets applications create, inspect, and manage repositories programmatically, so you can build your own code-hosting product on top instead of forking a monolith. Start as one node with a local directory, then separate compute from ledger nodes and replicate across three machines without pushing a repository again. 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.

Instead of babysitting three coding agents in three terminal windows, you describe the work once and let Mikka run the team: its Coordinator bot plans the job, dispatches runs to Claude Code, Codex, or Cursor on your connected machines, and only interrupts you when a decision is actually yours. Every piece of work lives in a thread with its own conversation, so dispatched runs appear as cards that report progress, ask questions, and flag blocks while you watch, reply, or stop them from the web, desktop, or phone. Unclear requests come back as choice cards before anything runs, and finished work moves through review and verification runs, back to a fix run when a check fails, then to acceptance where the pull request merges as you. Runs use leases and checkpoints, so a machine that drops mid-task hands the work to another connected computer with the same repository. You can create your own bots beside the Coordinator, each with its own chat, memory, and threads, and answers can arrive as charts, tables, stat cards, and task chips with sources cited. Daemons connect over SSH or from the desktop app, other agents drive Mikka through MCP, and Telegram or Slack messages reach the Coordinator too. 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.

WhatsApp is where your customers already are, and Waxum is the self-hosted gateway that lets your own software talk to them there without renting a per-message platform. One REST API fronts two providers at once: multi-device sessions paired by QR code or pairing code, and Meta's official Cloud API, sharing the same message routes and webhook events. Send text, media, locations, contact cards, polls, reactions, and broadcast lists, schedule messages for later, or run paced bulk blasts with deduplication, retries, and a dead-letter queue when a send fails. Voice and video calls ring, play audio, speak text in hundreds of voices, record the peer side, and stream media over WebSocket, while groups, contacts, blocking, privacy settings, and presence are all manageable per session. Signed webhooks with event filters and circuit breakers push receipts and replies into your app, and a failed delivery can be replayed from the queue instead of lost. The built-in ops console shows your whole session fleet, and each session opens a playground covering dozens of endpoints, so you can test a send from the browser before writing code. Sessions export and import between instances as zip files, and Prometheus metrics with health probes keep the fleet observable. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

You have a Proxmox cluster with spare capacity and people who keep asking you for a VM: ProxMate turns that cluster into an invite-only cloud where those people help themselves, inside limits you set. Invite links carry CPU, memory, and disk quotas, and tenants spin up VMs or LXC containers from an ISO, the Template Store, or twenty curated cloud images, pasting an SSH key to get a ready-to-SSH box in about a minute. Automatic placement picks the node with the image and the most free capacity, so nobody needs to learn your topology. Projects group machines with per-project usage totals and whole-project start and stop, and any VM can be shared with another tenant as viewer, operator, or manager, never with delete rights. Both a graphical noVNC console and a text console with copy and paste run in the browser, so access never needs open ports, and an optional per-VM browser IDE puts VS Code inside the guest. Backups run on schedules with rolling retention and one-click restore, snapshots are instant, power schedules stop dev boxes overnight, and time-boxed invites suspend access without deleting anything. Admins watch live per-VM sparklines, audit every action, and balance memory across nodes with live migration. 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.