Describe an app in one sentence and sandboxd, a self-hosted AI app builder, has a coding agent build it in an isolated sandbox and hand back a live preview URL. Start from a blank prompt, a React, Next.js, Express, or FastAPI starter, or an imported public Git repo. Refine it in one persistent chat beside Preview, Code, Terminal, and Git tabs. Pick OpenCode, Pi, or Claude Code. Point OpenCode or Pi at local or remote OpenAI-compatible and Anthropic-compatible endpoints, LiteLLM gateways included. A proxy injects provider keys on the wire, so nothing secret lands in a sandbox. Did the agent get it wrong? Every task is checkpointed first, so one click reverts the turn. Snapshots, forks, and named versions keep known-good points to branch from. Edit files in the browser, review file-by-file diffs, commit, and push. Secrets are write-only and encrypted per app. Idle sandboxes sleep and wake on the next request. One modest server can host per-user previews, coding playgrounds, or an agent platform. The app library installs 80+ curated open-source apps, such as Ghost, n8n, Gitea, and Grafana, each in its own sandbox. Every console action is also a /v1 HTTP call with an OpenAPI spec. Upgrades back up the database, health-check the new version, and roll back on failure. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

RustXNG gives your AI agents a private web search API that needs no keys, no accounts, and no per-search fees: ask once and get ranked results plus the text of the top pages, ready to ground an answer in real sources. A single request fans out to 180 search engines in parallel, drawn from 226 engine types covering general web, images, video, academic literature, package registries, and news, then merges what comes back with its own ranking that rewards term coverage and canonical-site authority instead of trusting any one engine's order. Engines that time out or keep answering nothing are suspended with backoff and rested automatically, so one dead source never stalls a task, and a repeated query returns from cache in about a hundredth of a second. Call it over HTTP with the familiar SearXNG-shaped JSON, CSV, RSS, or HTML formats, narrow results by domain, engine, or time range, or run the built-in MCP server and hand your agent native search and read_page tools with bounded, truncation-aware responses. Page fetches are hardened against untrusted URLs, an experimental local index can crawl your own corpus and merge it into results, and health, metrics, and per-engine status stay visible through the ops endpoints and the browser UI. 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.

Point every coding agent and chat client at one self-hosted gateway, then choose which model answers from a single browser console. Add providers by preset or custom OpenAI-compatible URL, or sign in with the subscriptions you already pay for. Routing groups bundle several models into one name and move to the next member when a quota runs out, so the calling agent never sees the error. Create a gateway key per client, cap each with daily token or cost budgets, restrict its models, and watch usage counted separately. The Usage pages chart tokens, cost, request counts, cache hit rate, and a yearly heatmap, split by model, provider, or agent, with CSV export. A Routing view replays how each request was routed, with timing for masking, connecting, and the vendor. Masking swaps passwords, API keys, and tokens for placeholders before a prompt leaves your server and restores them in the answer. The browser page is protected by a key you set at launch. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

OpenViking is a self-hosted context database that gives AI agents persistent, structured memory by organizing knowledge, skills, and session history into a hierarchical virtual filesystem accessible through the viking:// URI protocol. Instead of dumping everything into a flat vector store and hoping semantic search finds the right chunks, agents navigate their context with familiar commands like ls, tree, and find, locating exactly the information they need through deterministic paths combined with semantic search. Every resource is automatically processed into three layers: a 100-token L0 abstract for quick filtering, a 2,000-token L1 overview for content navigation, and the full L2 detail loaded only when confirmed necessary. This tiered approach cuts token consumption by 83 to 96 percent compared to conventional RAG while improving task completion rates by 15 to 49 percent on benchmark tests. The built-in memory self-iteration loop automatically analyzes task execution and user feedback, updating agent memory directories so the system continuously learns and improves. You can connect to any LLM provider, including Ollama for fully local inference, OpenAI, or compatible gateways. The Web Studio UI at the /studio endpoint provides visual browsing of the entire context filesystem, and the REST API on port 1933 supports programmatic access. Deploy via Docker, Kubernetes with the included Helm chart, or as a standalone service. 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.

MCPJungle puts every MCP server your AI tools use behind one endpoint, so Claude, Cursor, Codex, and your own agents all connect to a single gateway instead of a tangle of per-client configurations. Register remote HTTP servers and local STDIO servers once, and their tools, prompts, and resources become discoverable through the gateway in one place. Tool groups expose curated subsets at their own endpoints, so a coding agent sees only the tools its job needs instead of hundreds of irrelevant ones competing for context. The dashboard gives you a visual registry of servers, tools, and groups for everyday management, while the CLI and REST API cover scripting and automation. Development mode runs with zero authentication for a personal setup; enterprise mode adds per-client tokens that allow-list exactly which servers each client may reach. User accounts and manual server initialization give teams a governed shared deployment, with OpenTelemetry metrics showing what clients call and how the gateway performs. Upstream servers can authenticate with bearer tokens or OAuth, so credentials stay centralized in the gateway rather than copied into every client config. Registry data persists in a local file for solo use or in PostgreSQL when the gateway becomes shared team infrastructure. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MPL-2.0 licensed.

Hand your security alerts to a team of AI specialists and keep the final say on every action. Argus pulls events from cloud, identity, code, endpoint, and Kubernetes sources, runs fifteen detection rules, and turns matches into deduplicated findings enriched with CISA KEV and EPSS scores. Nine specialist agents hunt for evidence, build timelines, estimate blast radius, and propose fixes, but their only write path is a pending request that waits for human approval. Response playbooks can block an attacker or isolate a laptop through a catalog of 27 typed actions, starting in shadow mode with dry runs, caps, automatic rollback, and a tenant kill switch. Plant canary tokens aimed at rogue AI agents, replay eight simulated agent-driven attacks, and route OpenAI and Anthropic traffic through a gateway with model allowlists and budgets. Every step is written to a signed audit chain. A self-hosted command center with roles, TOTP sign-in, and Slack approvals fits small teams that have no SOC of their own. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

With 2,000 GitHub stars and 10 releases since its September 2025 launch, HeadlessX delivers a self-hosted browser automation platform that replaces Chromium-based scraping with Camoufox — a Firefox fork performing kernel-level fingerprint spoofing to achieve 0% detection across Cloudflare, DataDome, PerimeterX, and other anti-bot systems where Puppeteer and Playwright regularly fail. The web dashboard provides workspace-based job organization with a visual interface for configuring scrape targets, managing browser profiles, monitoring queue status, and viewing extracted results in real time. The protected REST API accepts requests with API key authentication for programmatic access, supporting HTML extraction, screenshot capture, PDF generation, and structured data parsing with configurable stealth parameters. Profile-based scraping maintains persistent browser contexts with cookie jars, localStorage, and fingerprint configurations that survive between requests — reducing cold-start latency from 25 seconds to under 2 seconds on subsequent requests. Queue-backed workflows enable batch processing of URLs with configurable concurrency, retry logic, and webhook notifications on completion. The Google AI Search integration provides AI-assisted web research workflows through dedicated endpoints. Remote MCP support exposes automation capabilities as tool endpoints for AI agent integration. Deploy via the official CLI with `headlessx init` and `headlessx start` commands, scaffolding a Docker Compose stack with Caddy reverse proxy for automatic HTTPS. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Super Agent Party gives you an AI companion with a face, a memory, and a to-do list, all running on your own server. The home screen starts a chat, summons a companion, creates a role or agent, queues a task, opens an AI browser, or adds an extension. Characters can appear as VRM 3D avatars, and character cards with long-term memory let several personas talk together in one group chat. The Task Center hands longer jobs to an agent that works in the background, calls tools through MCP and Agent Skills, and records every step in an execution timeline. A built-in browser lets agents open pages and act on them, and desktop vision tools can drive the mouse, keyboard, and terminal. One-click bot deployment connects an agent to Telegram, Discord, Slack, QQ, WeChat, Feishu, and DingTalk, and live-stream bots cover Bilibili, YouTube, and Twitch. The developer page serves an OpenAI-style endpoint, so other apps can call an agent as if it were a model. Extensions open in a sidebar or their own window. 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.

Guaardvark is a complete AI studio that lives on your own GPU, so chat, retrieval, agents and media generation happen on hardware you own instead of a metered cloud account. Upload documents and question them through local retrieval that fuses keyword and vector search, reranks candidates, and cites the passages behind each answer, with the chunks and scores visible for inspection. Plan mode turns a request into an approve-or-edit plan card before tools run, and approved steps execute one at a time with a pause on failure. Screen agents receive a desktop of their own, see it through a vision model, and operate a real browser while the operator watches from any page. Coding swarms split a task across parallel agents in separate git worktrees and merge results back in dependency order. The generator bench covers thirteen local video models, image models for generation, editing, inpainting and upscaling, full songs with vocals, sound effects and three text-to-speech engines, with consent required before any voice clone or public post. Character, environment and prop LoRAs train from a few reference photos and keep the same character consistent across images and films. A GPU orchestrator with a live VRAM budget schedules renders, chat and training so jobs stop competing for the card. 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 pull a team's skills, MCP servers, hooks, prompts, and agent files out of scattered Git repositories, Observal keeps them in one reviewed registry and traces real sessions to show which ones help. Build an agent visually, with a live config preview for every harness. Submissions wait in an admin review queue, where reviewers inspect side-by-side version diffs before approving a release. Teamspaces give each team its own namespace, and approved releases can be forked into independent drafts. One pull command then writes the right config for Claude Code, Cursor, Kiro, Codex, Copilot, Pi, OpenCode, Antigravity, or Goose. Every approved resource is discoverable through the open Agentic Resource Discovery standard, and a running agent can hand part of its task to another approved agent or a remote Agent2Agent service. Hooks record each session as a trace of prompts, tool calls, MCP round trips, tokens, and errors, with turn-by-turn replay for debugging and audits. Insight reports, generated through providers such as Anthropic, OpenAI, or Ollama, list what is working, what is hindering, and quick wins. A dashboard follows adoption, cost, and department usage, a leaderboard ranks agents by downloads, and audit logs, SAML single sign-on, and SCIM provisioning are included. 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 team building with coding agents has the same drawer problem: the best Skill, MCP server, or agent file lives in one person's repo, undocumented, while three colleagues rebuild it from scratch. Agentic Resource Registry is the internal catalog that ends the duplication, a self-hosted registry where your organization publishes Skills, MCP servers, hooks, prompts, agent files, and sandboxes in one searchable place. Browse by teamspace, publisher, category, or harness and see at a glance what each resource supports and how often it has been pulled. Authors build agent files in a four-step builder with a live summary and draft readiness checks, then submit them into a review queue where admins inspect the full prompt and version diffs before approving. Forking an approved release creates your own draft with pinned components, and reviewers can compare it against upstream line by line. The companion CLI installs any approved resource into Claude Code, Cursor, Codex, Kiro, Copilot, OpenCode, Pi, and more, generating the correct config for each harness automatically. Approved resources are also discoverable over the open ARD standard, so a running agent can hand part of its task to another approved agent or a remote A2A service. 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.

Watch every coding agent you run, on every machine, from one live browser dashboard. Sessions appear as cards, a sortable list, or a graph, grouped by machine or project, each showing its plan and current step, subagents, model and effort, context fill, git branch, and cost. Anything waiting on you, like a permission, a question, or a plan to approve, rises to a Needs you panel with the longest wait first, and you can answer it from the page or in the terminal. A Problems panel flags sessions that are stuck, failed, or colliding on the same file. The Plans board sorts every plan step into pending, in progress, review, blocked, and done. A session page adds a timeline of loops and subagents, changed files, and token and context charts. Spend is broken down by model and machine, and plan windows show how close each provider is to its limit. History keeps ended sessions by day, a wall view suits a spare screen, and browser alerts fire when a session needs 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.

Mention an AI agent in a channel the way you would page a teammate, and it picks up the surrounding conversation, claims the task, does the work in its own persistent workspace, and reports back in the thread. OpenTag is a self-hosted, Slack-style workspace where people and coding agents from Claude Code, Codex, GitHub Copilot, and other runtimes work as one team across channels, threads, DMs, and a shared task board. Each agent keeps its own memory file, runtime session, permissions, and activity history, can be started, stopped, put to sleep, woken, or resumed, and can delegate pieces of a job to other agents or schedule reminders that wake it later. Teams watch live progress through per-message activity that expands to show thinking, status, and tool calls, while full-text search keeps every decision findable where it was made. The control plane runs from Docker Compose on your server, connected machines run the daemon that hosts the agent CLIs, and attachments, databases, and workspaces stay on infrastructure you control. Multi-workspace accounts, scoped permissions for agents and members, and per-channel task numbering keep larger teams organized without surrendering their context to someone else's cloud. 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.

The most comprehensive open-source control plane for Hermes Agent — a full workspace combining AI chat, visual workflows, multi-agent orchestration, coding agent management, and platform channel integration in one self-hosted dashboard. Real-time chat streaming over Socket.IO connects to any OpenAI-compatible backend including Ollama, OpenAI, Anthropic, and custom endpoints with multi-session management, tool call expansion, inline file previews for HTML, PDF, DOCX, images, and source code, plus profile-scoped uploads and workspace attachments. The visual workflow builder provides a Vue Flow canvas for constructing DAG-structured pipelines with directed edges, conditional routes, approval gates, loops, and live execution with per-node status updates. Platform channel integration configures Telegram, Discord, Slack, WhatsApp, Matrix, Feishu, DingTalk, QQBot, WeChat, and WeCom bots from one page with credential management and per-platform behavior settings. Multi-agent group chat rooms enable real-time messaging with @mention routing, automatic context compression, and SQLite message persistence. The coding agent panel installs, launches, and monitors Claude Code and Codex with built-in terminal, session history, and file diffs. Usage analytics track token consumption, estimated costs, cache hit rates, and 30-day daily trends with model distribution charts. Kanban boards plan and track agent work alongside cron job scheduling for recurring tasks. Deploy via Docker, npm CLI, or desktop installer for Windows, macOS, and Linux. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Most AI chats forget you by tomorrow; Hivekeep keeps a team of agents on your own server that remember everything, work together, and answer wherever you already talk. Each agent has a name, a role, and one continuous conversation that never resets, backed by long-term memory it can search. Agents hand tasks to one another, spawn helpers, and run scheduled jobs. Queenie, the built-in setup agent, connects providers and creates your first teammates through plain chat. They reach you on Telegram, Slack, Discord, WhatsApp, Signal, Matrix, or the installable web app, and a channel can move from one agent to another mid-conversation. Agents write custom tools, build mini-apps that open beside the chat, edit files in a workspace with a code editor, and add plugins from a marketplace. Credentials go straight into an AES-256-GCM vault that the model never reads. A context viewer and a token-usage dashboard show what each prompt contains and what each agent, model, and day costs. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Agents with tool use, deep research, a document editor, an IMAP/SMTP email client with AI triage, notes, tasks, and a CalDAV-synced calendar - Odysseus bundles all of it into one open-source, self-hosted AI workspace. It runs local models through Ollama, vLLM, or llama.cpp and cloud APIs like OpenAI and OpenRouter, with a hardware-aware Cookbook that scans your machine and recommends quantized models that fit. Persistent memory uses ChromaDB with hybrid vector-plus-keyword retrieval, web search runs through a bundled SearXNG instance, and agents can use MCP servers, files, and shell access with safety controls, plus custom skills and scheduled agent tasks. A blind Compare mode runs side-by-side model duels with identities hidden and accumulates Elo-style ratings from your votes, so model selection is based on your actual workloads rather than leaderboard claims. Deep research mode - adapted from the Tongyi DeepResearch approach - reads sources through SearXNG and produces cited reports, while the email client tags, summarizes, sets reminders, and drafts replies locally rather than through a third-party mail AI. The writing-first document editor adds AI edits, Markdown and HTML support, and version history. The stack is Python 3.11 with FastAPI, SQLite for state, and a vanilla JS frontend, licensed AGPL-3.0 with zero telemetry. Because agents can read email and execute commands, keep authentication enabled and never expose it as a public unauthenticated service.

OpenAgentCore puts Codex, Claude Code, and MiniMax Code behind one self-hosted Agents API, so your applications start agent sessions through the official OpenAI SDK while the work runs on machines you control. Set OPENAI_BASE_URL to your installation and issue a Project API key, and the same client code that talks to OpenAI now talks to your Core. Each session runs a native harness, exactly as it runs on a developer's machine, with the model provider, model ID, and protocol you configure per harness or per session. Choose where sessions execute: microsandbox nodes you operate, Docker, E2B, or a self-hosted workstation, VM, or container connected by the oac-daemon with a credential scoped to one environment. Idle microsandbox sessions release their compute and resume on a compatible node with workspace files intact, so long investigations do not rent a machine while they sleep. The Web console walks operators through setup, tracks sessions and turns, manages models and credentials, and issues keys, while the separate Core API serves operators and the Agents API serves applications. Durable execution state in PostgreSQL keeps events, results, and schedules recoverable across restarts. Sandbox providers, workspace filesystems, harnesses, and model providers all plug in through defined protocols, so every layer can be replaced on its own. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Claiming 60-80% token cost reduction compared to flat single-model deployments and backed by 6,500+ GitHub stars, OpenSquilla delivers an intelligent AI agent runtime where a local ML classifier evaluates every turn on message length, code blocks, keyword patterns, and semantic embeddings before routing it to the optimal model tier from C0 through C3. The pluggable provider layer connects natively to TokenRhythm, OpenRouter, OpenAI, Anthropic, Ollama, DeepSeek, Gemini, DashScope, Moonshot, Mistral, Groq, Zhipu, SiliconFlow, vLLM, LM Studio, and additional compatible backends with primary-plus-fallback selection. The four-tier cognitive memory architecture spans working, episodic, semantic, and raw layers with vector-semantic and BM25 retrieval powered by on-device ONNX embeddings that never leave your infrastructure. Security isolation operates at the syscall level via Bubblewrap on Linux and Seatbelt on macOS, complemented by policy-based execution controls and prompt injection protections. The unified TurnRunner executes identically across the Vue-based control console Web UI, terminal CLI, and chat channel integrations including Slack and Discord, ensuring consistent tool dispatch, retry logic, and decision logging regardless of entry point. Built-in skills cover deep research, multi-search-engine queries, document generation for DOCX, PPTX, XLSX, and PDF formats, GitHub integration, cron scheduling, and bounded subagent delegation. Per-agent workspaces with durable session storage provide transcript replay, context state management, and per-call cost tracking with automatic quota enforcement. 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.