OpenTranscribe is a self-hosted transcription platform that converts your audio and video library into searchable, speaker-labeled transcripts, then lets you summarize and chat with those recordings. Upload files up to 15 GB, record from the browser microphone, or paste a link from YouTube and a thousand other sites, and playlists queue up for batch processing automatically. WhisperX produces word-level timestamps in over 100 languages, while the diarization engine works out who spoke when, corrects turn boundaries, and recognizes the same voice again across different recordings. Speaker profiles accumulate a voice fingerprint, speaking-time analytics, and merged identities, so a regular guest is named correctly in every episode without relabeling. Point it at an LLM and recordings gain structured summaries with action items, suggested tags and collections, and a chat interface that answers questions about your transcripts with citations, grouped into projects that pin their source recordings and standing instructions. Leave the model unconfigured and transcription, diarization, full-text and semantic search, and redaction still run entirely on local models. Watch folders, S3 buckets, and network shares import new media on a schedule with three-layer deduplication and automatic stitching of split recordings, and everything exports as subtitles, Word or PDF documents with citeable numbered lines, or Markdown. 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.

Scoring 64.4 on the WebBench benchmark — state-of-the-art among browser automation platforms — Skyvern replaces brittle XPath-based scripts with Vision LLM reasoning that adapts when websites change their layouts. The platform extends Playwright with AI-powered page methods including page.act(), page.extract(), and page.validate() that accept natural language prompts while still supporting traditional CSS selectors as fallback. The drag-and-drop Workflow Studio offers 17+ block types including navigation, extraction, login, loops, conditionals, code blocks, file download, and file upload — enabling non-technical users to build complex multi-step automations without writing code. Self-hosted deployments support bring-your-own-LLM with OpenAI, Anthropic, Gemini, and Ollama, while the multi-engine architecture allows swapping between Skyvern 2.0, OpenAI CUA, Anthropic CUA, or UI-TARS per task with a single parameter. Built-in infrastructure handles persistent browser sessions preserving cookies and localStorage across runs, automatic CAPTCHA solving for reCAPTCHA and hCaptcha, anti-bot bypass for Cloudflare and DataDome, residential proxy rotation across 20+ countries, and a credential vault integrating with Bitwarden and 1Password for secure 2FA management. Real-time session livestreaming via WebRTC enables visual debugging, while step-by-step action logs with screenshots and full LLM diagnostic traces provide production observability. The MCP server integration exposes Skyvern as a tool for Claude, Cursor, Windsurf, and any MCP-compatible AI agent. Connect to 6,000+ apps through Zapier, Make.com, or self-hosted N8N workflows. Deploy via Docker Compose or pip install with a two-command setup. 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.

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.

With 2,600+ GitHub stars since its May 2026 open-source launch, Knowhere solves the last-mile problem of document intelligence for AI systems — transforming complex unstructured PDFs, reports, and multi-page documents into structured JSON chunks that LLMs can consume without hallucination. The platform processes documents through an AI-native parsing pipeline that handles 20+ page documents with deep hierarchies, intricate tables, and multimodal content including images with OCR, achieving 95% precision in information extraction while reducing token costs by 50% compared to raw document ingestion. The knowledge tree architecture maintains historical context across multiple documents, enabling cross-document graph navigation for agentic retrieval that goes beyond simple chunk-based RAG. Built on Python 3.11+ with MinerU as the default PDF parser, the backend API runs alongside async workers that process document ingestion, graph construction, and embedding generation. The self-hosted Docker Compose stack packages the API server, processing workers, and Next.js dashboard for managing API keys, webhooks, and document-processing jobs, backed by PostgreSQL and Redis. Both Python and Node.js SDKs provide programmatic access for integration into existing AI pipelines and agent frameworks. LLM providers include DeepSeek and Alibaba Cloud DashScope with configurable key rotation for rate-limit management. Deploy 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.

AIHOT is a self-hosted engine for running your own AI news site: it watches the sources you choose, scores every story twice, and publishes only the coverage worth reading. Point it at RSS feeds, web list pages, JSON APIs, X accounts, WeChat public accounts, or content pushed in by scripts, polled on schedules you control. After deduplication and pre-screening, every story faces two independent scoring passes under the same rubric, so one generous run cannot wave a weak story through. Coverage of the same story clusters into one event across the last two weeks, so forty outlets on one launch read as a single developing story. Hotness is measured per event: each source counts once per 48 hours and its weight halves daily, so repeat posting cannot inflate the ranking. Digests compile on schedule: a daily edition at 08:00, a weekly review on Mondays, and a monthly rollup on the first, assembled by rules without model calls. Readers browse topic pages by company, research direction, and content form, search titles, summaries, and full text, and keep local favorites with export and import. From the admin backend you can test sources, correct scores by hand, calibrate selection against labeled samples, and swap the model behind each pipeline step. Output flows through RSS, a public API, MCP, agent Markdown, and llms.txt, so humans and AI agents can subscribe. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Ever Works transforms a single text prompt into a fully researched, written, coded, and deployed website through autonomous AI agents that operate on a configurable 15-step generation pipeline. Unlike one-shot builders, it continuously improves generated content on a schedule, keeping sites fresh without manual intervention. The platform integrates with six AI providers (OpenAI, Anthropic, Google, Groq, Mistral, Ollama) via gateways like OpenRouter and Vercel AI Gateway, while its plugin system connects ten search providers including Tavily, Brave, Exa, SerpAPI, and Perplexity for deep autonomous research. Every generation is Git-native — code and content live in your own GitHub repositories as commits, providing full version control, portability, and zero vendor lock-in. The monorepo ships a NestJS 11 REST API with JWT authentication, a Next.js 16 web dashboard for managing works, a CLI for terminal operations, an MCP server for AI agent integration, and admin tooling. Content management supports items, categories, tags, and collections, while deployment targets include Vercel and Kubernetes. Background processing runs on Trigger.dev with BullMQ scheduling, and monitoring leverages Sentry plus PostHog for error tracking and product analytics. The platform supports OAuth via GitHub and Google, subscription billing, email notifications, and AI chat conversations for natural-language interaction with your generated works. 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.

Ask your assistant what you spent on groceries last month, and let it answer from your actual bank accounts instead of a spreadsheet you never update. BankMCP is a small server you run yourself that connects your European banks to Claude, ChatGPT, Cursor, or any MCP client, so questions about balances and spending get real answers. Ask for a balance and it fetches the booked and available figures live; ask where the money went and it searches transactions across every connected account by merchant or invoice number; ask for the month and it totals money in and out while leaving transfers between your own accounts out of the picture. Background watches keep an eye out between conversations: get a webhook when a balance crosses a line, a large debit lands, a named payment arrives, or an expected payment has not shown up by its date. Ready-made prompts cover connecting a bank, building a budget, scanning savings, auditing subscriptions, and spotting unusual transactions. The server is deliberately boring about your data: it keeps consents, account labels, watches, and your login, fetches balances and transactions only when asked, and stores none of them. It is read-only by design, with no payment tools to abuse. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Atomic transforms collections of personal Markdown notes into an AI-augmented knowledge graph by breaking documents down into semantically linked conceptual units. Users capture thoughts, research excerpts, and web clippings as modular notes that the system automatically chunks, embeds, and categorizes with hierarchical tags. An interactive force-directed canvas maps out the entire knowledge base, positioning related ideas closer together based on embedding cosine similarity rather than manual folder hierarchies. Built-in vector search queries across text embeddings directly within SQLite, retrieving precise contextual snippets in response to natural language prompts. The automated wiki engine synthesizes disparate notes into comprehensive reference articles equipped with verifiable inline citations back to original source materials. Integrated conversational agents leverage retrieved knowledge context to answer research questions, surface contradictions across past entries, and draft structured daily briefings. External AI tools can read, search, and update stored knowledge through an embedded Model Context Protocol server 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.

Trusted by thousands of organizations with over 30 million monthly downloads and 20,000+ GitHub stars, MLflow is the largest open-source AI engineering platform providing end-to-end lifecycle management for traditional ML models, LLMs, and AI agents. The OpenTelemetry-based tracing system captures complete request flows through any LLM provider or agent framework — including OpenAI, LangChain, DSPy, Vercel AI, PydanticAI, and smolagents — with one-line auto-instrumentation that tracks inputs, outputs, token usage, and costs at every intermediate step. MLflow's evaluation engine offers 50+ built-in metrics and LLM judges for systematic quality assessment, detecting issues across correctness, latency, adherence, relevance, and safety dimensions before code reaches production. The Prompt Registry versions, tests, and deploys prompts with full lineage tracking while automated optimization algorithms improve prompt performance using evaluation feedback. The AI Gateway provides a unified API endpoint for all LLM providers, enforcing rate limits, cost controls, and access policies across the organization. MLflow 3.0 introduces the LoggedModel abstraction linking traces, metrics, and prompts to specific model versions across Python, TypeScript, Java, and R SDKs. The model registry manages deployment workflows with automated quality gates, while experiment tracking records parameters, metrics, and artifacts across training runs. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache License 2.0 licensed.

Marimo is a reactive Python notebook that treats cells like spreadsheet formulas: change one cell or interact with a UI widget and every dependent cell automatically re-executes, eliminating the hidden state bugs that make traditional notebooks unreliable. Backed by over 22,000 GitHub stars, notebooks are stored as pure Python files with PEP 723 inline metadata, making them Git-diffable, importable as modules, executable as CLI scripts with parameterized arguments, and testable with PyTest. Built-in SQL cells query Polars, Pandas, PyArrow, DuckDB, SQLite, PostgreSQL, and MySQL databases, with results automatically flowing into the reactive dependency graph. The AI-native editor provides GitHub Copilot autocomplete, context-aware assistants that access live runtime variables, inline code edits powered by configurable models from OpenAI, Anthropic, or local Ollama instances, and a pair mode that lets external AI agents connect over WebSocket. Notebooks become read-only interactive web applications with marimo run, collaborative authoring environments with marimo edit, or embedded flows inside existing FastAPI applications through ASGI middleware. Gallery mode serves multiple notebooks from a single instance with an auto-generated index page. The Docker image ships with SQL support, token-based authentication, health check endpoints at /health and /api/status, and configurable WebSocket or SSE kernel transport for proxy compatibility. 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.

Xalgorix delivers AI-driven penetration testing that proves vulnerabilities exist rather than guessing, pairing an autonomous LLM agent with an independent exploit verifier that re-exploits every finding before it reaches your report. The 22-phase methodology mirrors how a skilled human pentester works through an engagement, covering reconnaissance, subdomain enumeration, port scanning, web crawling, parameter fuzzing, SQL injection testing, XSS detection, authentication bypass, API testing, and more. Select all phases or pick specific ones per target. The platform ships with 85+ offensive security tools preinstalled (nmap, nuclei, httpx, subfinder, katana, ffuf, gobuster, sqlmap, masscan, dalfox, feroxbuster) plus 14 built-in agent tools for browser automation, terminal execution, and note-taking. Connect your own LLM provider (OpenAI, Anthropic, DeepSeek, Gemini, Groq, Ollama, or MiniMax) so no scan data or target information leaves your infrastructure. The web dashboard on port 9137 provides live WebSocket telemetry showing tool calls, agent reasoning, and findings as scans progress. A findings index with CVSS scoring and severity filters organizes results, and branded PDF reports include your company name and logo. Wildcard and multi-target scans handle red team attack-surface mapping, while source-code scanning audits repositories directly without requiring a deployed target. A REST API enables scripted scans and SIEM pipeline integration, and webhooks push notifications to Discord or Telegram. 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.

T3 Code puts every coding agent you already pay for behind one control surface, so Claude Code, Codex, Cursor, Grok Build, OpenCode, and Antigravity all run on your server while you direct them from your phone, browser, or desktop. Start a thread, choose the provider and model that fit the job, and watch the agent work in a shared conversation instead of a terminal you have to babysit. Permission modes decide when an agent must stop and ask, and its approvals, questions, and plan reviews arrive in the same thread wherever you happen to be connected from. When the work lands, review it file by file in the built-in diff view, then commit, push, and open pull requests through the Git and GitHub integrations without leaving the app. A real terminal sits beside the chat, and projects, settings, and scheduled tasks stay attached to the machine where the code actually runs. Install it as a background service and reach it from anywhere through T3 Connect, a Tailscale or LAN pairing, or the hosted web client, with a pairing code guarding local access. Outside agents can also drive threads over MCP, launching, messaging, and waiting on work across machines you have linked together. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

The fastest tool on the public Martian Code Review Bench at 77 seconds per PR with competitive F1 scores — while every higher-scoring competitor takes 5 to 14 times longer — Mira is the fully open-source AI code reviewer that ships the entire feature surface other tools gate behind per-seat subscriptions: indexed PR reviews with full-repo context, streaming walkthroughs posted within seconds, inline vulnerability scanning from hourly OSV.dev polls, org-wide package inventory answering "which repos use [email protected]?" in one query, interactive dependency graphs with blast-radius SVG rendering via ReactFlow, a learning loop that synthesizes rules from rejected comments and human review patterns on merged PRs, and cost telemetry showing actual spend per repo and per model. Unlike CodeRabbit, Greptile, and GitHub Copilot code review which are SaaS-only with per-seat pricing, Mira runs as a single Docker image on your infrastructure with your LLM key through OpenRouter (fronting Anthropic, OpenAI, Google Gemini, DeepSeek) or direct to Ollama and vLLM for fully air-gapped deployments. The engine indexes Python, TypeScript, Go, Rust, Java, C/C++, C#, Swift, Kotlin, Scala, and PHP with per-file symbol extraction, enabling multi-file reasoning across diffs rather than isolated line-by-line review. Integrates with GitHub, GitLab, and Forgejo via webhook with full feature parity. 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.

OpenGTM turns a market, a list, or a workbook into enriched, researched, ready-to-contact prospects on servers you control. Give it a market and it sources companies and people; hand it a CSV or a Clay table and it maps the columns into a shared record library. Enrichment columns run as waterfalls that try providers in cost order until a field is filled, and before anything spends a cent you see best and worst case estimates, with a workbook budget ceiling that halts paid calls on its own. Research columns send an agent to answer the hard question for each row, and the answer keeps citations to the pages it actually fetched, so a human can audit the claim before outreach. Live audiences segment your records and watch for hiring and buying signals, refreshing on schedules and firing automations when a prospect heats up. Finished rows flow out to HubSpot, Salesforce, Slack, Sheets, Airtable, Instantly, Smartlead, webhooks, or a warehouse, while every cell keeps a trace of provider attempts, latency, and cost. Your own agents get the same workspace through a REST API, webhooks, and an MCP server with scoped read and write tokens, and a seeded zero-key demo means the team can explore the whole loop before adding a single provider key under Settings. 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.

DramaClaw is a self-hosted production line that carries a manuscript or screenplay all the way to a finished AI drama, handling story structure, characters, storyboards, voices, and final cuts inside one workspace. Drop in a novel manuscript or a Fountain screenplay and it builds episodes, characters, and scenes for you, then plans multi-episode arcs and writes scripts in adaptive, literal, or staged modes with review-and-repair loops. The shared asset library keeps every character, scene, prop, and voice organized once, so faces stay consistent across episodes while portraits and per-episode variants handle costume and age changes. Storyboards and first frames are generated beat by beat, emotion-aware voice-over is synthesized per character, and episodes are assembled and exported with subtitles and the full asset pack. For finer control, the XiaHua canvas gives you 18 node types for images, video, audio, scripts, and 3D sets, with a wall of 45 short-drama styles and tools for relighting, outpainting, upscaling, and camera moves. Xia Director, the built-in agent, turns a plain-language brief into a node graph, runs it with progress tracking, and can promote winning results from the canvas back into the series. You pick the models: paste an official gateway key, or route Seedance, Kling, Hailuo, Sora, ComfyUI, and other channels through the bundled gateway, while external agents drive the same pipeline over MCP. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Elastic License 2.0 licensed.

Previously known as Hoarder and now holding 28,000+ GitHub stars, Karakeep is the most popular open-source bookmark-everything application — combining AI-powered automatic tagging with full-text search, page archival, and cross-platform access for digital content hoarders who refuse to let valuable links disappear. The Next.js frontend with tRPC communication delivers a responsive interface for saving links, notes, images, and PDFs, while Puppeteer crawls bookmarked pages to fetch titles, descriptions, and images automatically. LLM-based auto-tagging supports OpenAI, Anthropic, or local models via Ollama for privacy-first deployments that never send data to external services. Meilisearch powers full-text and semantic search across all stored content including OCR-extracted text from images. A rule-based automation engine triggers custom actions based on bookmark properties — automatically sorting, tagging, or archiving content matching defined conditions. Full page archival via Monolith preserves complete page snapshots against link rot, while yt-dlp integration archives videos from YouTube and other platforms. RSS feed ingestion automatically captures new articles from subscribed sources. Collaborative lists enable teams to build shared bookmark collections, with per-list permissions and real-time sync. Native iOS and Android apps, Chrome and Firefox extensions, and browser bookmark sync via Floccus ensure capture from any device. Importers migrate data from Chrome, Pocket, Linkwarden, Omnivore, and Tab Session Manager. 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.

Kokoro-FastAPI turns text into natural-sounding speech across eight languages by serving the 82-million-parameter Kokoro-82M model through an OpenAI-compatible REST API, so any existing OpenAI SDK client can generate audio by just changing the base URL. With over 5,300 GitHub stars since December 2024, the fully Dockerized FastAPI server covers American English, British English, Spanish, French, Hindi, Italian, Japanese, Brazilian Portuguese, and Mandarin Chinese with language-specific phoneme processing. Inline voice mixing blends multiple profiles using weighted ratios like af_bella(2)+af_heart(1), automatically normalizing weights and caching combined voicepacks as PyTorch tensor files for reuse. Audio streams in real time over HTTP with configurable chunk sizes, or generates complete files in MP3, WAV, OPUS, FLAC, AAC, or PCM formats with speed control from 0.25x to 4.0x. Per-word timestamped captions with speaker-tagged voice labels enable subtitle generation for podcasts, audiobooks, and accessibility workflows. Pre-built Docker images support NVIDIA GPU acceleration via CUDA, experimental AMD GPU inference via ROCm, and CPU-only deployment on linux/amd64 and linux/arm64 architectures, with Apple Silicon MPS support available through direct UV execution. The integrated web interface at port 8880 provides browser-based speech generation, while the Swagger UI at /docs exposes the full API reference. Debug endpoints report system statistics for monitoring inference load. 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.

Information overload is the default state of the modern internet, and Glean exists to tame it by pulling every blog, newsletter, and news source you follow into a single, organized reading experience with an interface that prioritizes clarity over clutter. Subscribe to RSS and Atom feeds, import your existing subscriptions via OPML, and organize everything into multi-level folders with a flexible tagging system that lets you slice your reading list by topic, priority, or mood. A three-pane layout presents your feed list, article summaries, and full content side by side, with one-click actions for marking articles as read, saving them for later, bookmarking favorites, or archiving what you have already processed. Background workers fetch new articles every 15 minutes so your feeds stay current without manual refreshes. The preference learning engine (backed by Milvus vector embeddings) scores incoming articles against your reading history and surfaces a personalized Smart Feed of content you are most likely to care about, reducing the time spent scrolling past irrelevant posts. A built-in MCP server lets AI assistants like Claude Desktop search and interact with your feed library directly. The admin dashboard on a separate port provides user management, system monitoring, and configuration controls for multi-user deployments. Bookmark external URLs from any browser with the Chrome extension, or access your feeds on mobile through the Progressive Web App. 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.