DeepSeek Harness gained over 60,000 GitHub stars within hours of its August 2026 launch, establishing itself as the first fully modular open-source agent runtime where literally every component is a swappable plugin. Built on the Cordis framework—a programming paradigm for spatiotemporal composability—dsh decomposes the entire agent stack into independently replaceable pieces: model adapters for DeepSeek, Anthropic, OpenAI, AWS Bedrock, Azure, and Google Gemini; tool registries covering bash execution, file system operations, web search, subagent delegation, and todo management; plus session stores, sandboxes, approval policies, orchestration loops, and the user interface itself. Four operating modes serve different workflows: Standard provides the full toolset, Code mode uses model-generated code to compose multi-round tool calls, Minimal strips down to a shell and editor for benchmarking, and Creator mode lets developers inspect the running runtime and test Cordis plugins in memory. The kernel handles plugin mounting, unmounting, and dependency resolution while typed events and services coordinate between components. Profiles and bundles allow the same codebase to produce entirely different products—a terminal coding agent, a browser-based workspace, a headless automation service, or an ACP/JSON-RPC endpoint—by swapping YAML configuration layers. Session history is stored as an append-only event stream for full trajectory replay, and project-level hooks on agent lifecycle events enable fine-grained behavioral customization. MCP client integration connects to external tool servers, while Agent Client Protocol enables programmatic orchestration. 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 83,000+ GitHub stars and $18.8M in Series A funding, OpenHands delivers the leading open-source platform for AI coding agents that scored 68.4% on SWE-bench Verified with Claude Opus 4.6, outperforming Devin 2.0's publicly reported 45.8%. The Agent Canvas web UI organizes work into persistent conversations where agents edit files, run shell commands, browse the web, and execute multi-step development tasks inside isolated Docker sandbox containers. The observe-plan-act loop drives agent behavior: the Python controller manages LLM abstraction via LiteLLM routing to 100+ providers including OpenAI, Anthropic, Google, DeepSeek, Qwen, Llama, and local Ollama models. Built-in skills for code review, Docker management, PRD generation, repo-rules enforcement, release notes, and test running attach to conversations automatically via auto-discovery or trigger-based activation. The Automations system schedules recurring agent tasks with configurable templates for CI workflows, dependency updates, and documentation generation. MCP server integration enables agents to access external tools and data sources. The REST API powers an OpenAI-compatible endpoint for connecting agents to chat UIs, IDEs, and voice platforms. GitHub, GitLab, Slack, and Jira integrations enable pull request reviews, issue resolution, and team notifications. The SDK provides Python and REST APIs for embedding agents in custom tools with local or cloud execution, custom agent behaviors, and Kubernetes deployment. On RepoCloud, deploy OpenHands on a dedicated VPS with Docker socket access, persistent project storage, root SSH access, and complete control over your AI development infrastructure, all under the MIT license.

With over 82,000 GitHub stars and 700,000+ downloads, LobeHub has evolved from its origins as LobeChat into a comprehensive multi-agent AI collaboration platform where humans and autonomous agent teams co-evolve. The platform's Agent Harness architecture functions as an operating system between AI models and applications, handling prompt presets, tool orchestration, lifecycle hooks, planning, filesystem access, and sub-agent management across 25+ model providers including OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Mistral, Groq, AWS Bedrock, Azure OpenAI, and local models through Ollama. Agent Groups enable sophisticated collaboration with sequential, parallel, iterative, and debate orchestration modes, allowing multiple specialized agents to tackle complex workflows simultaneously. The Agent Builder creates production-ready agents from natural language descriptions with auto-configuration, drawing from a marketplace of 505+ pre-built agents and 10,000+ MCP-compatible skills and plugins. Pages provide collaborative document editing with multi-agent co-authoring, while Schedules automate agent runs around the clock without human supervision. The knowledge base leverages PostgreSQL with pgvector for RAG-powered retrieval, and Personal Memory gives agents transparent, editable context that evolves through continual learning. The full self-hosted stack deploys via Docker Compose with PostgreSQL, Redis, RustFS for S3-compatible storage, and SearXNG for private web search, all configurable through environment variables. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. LobeHub Community licensed.

With over 2,100 GitHub stars in its first month and benchmarked at 30-75% lower cost than Claude Managed Agents on enterprise task suites, TrueForge is the open-source agent harness that provides the complete runtime layer for turning any LLM into a working production agent on your own infrastructure. The TypeScript server runs the full execution loop — streaming every step, routing tool calls through MCP servers with centralized header-auth and in-chat OAuth, delegating parallelizable work to isolated subagents, and pausing for human approval on sensitive actions. Context engineering keeps token costs low: deferred tool-schema loading delays MCP schemas until invoked, large-result offloading moves oversized outputs to files, Code Mode processes structured data through sandboxed execution, and automatic compaction summarizes older history at a configurable 50,000-token threshold while preserving the full transcript. The sandbox-as-a-tool architecture provisions isolated Daytona environments only when code execution is required, allowing one server to run many concurrent agents without idle overhead. Agents are configured from shipped YAML catalogs of models, MCP servers, git-backed SKILL.md instruction packs, and sandbox providers, then saved to an Agents Library accessible via the chat UI, TypeScript SDK, or embeddable React UI SDK. Run locally with SQLite via a single npx command, or deploy for teams with Docker Compose or Helm using Postgres and Redis with OIDC authentication. 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.

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.

With over 110,000 GitHub stars and one of the fastest-growing open-source communities in the self-hosted space, Immich delivers a Google Photos-grade experience entirely on your own hardware. The platform handles automatic background backup from Android and iOS devices, deduplication, and support for RAW formats, LivePhotos, and MotionPhotos. Its machine learning pipeline runs facial recognition and clustering locally on your server, enabling you to group photos by person without sending a single image to the cloud. CLIP-based semantic search lets you find images by describing their content in natural language, while metadata-driven search covers EXIF data, dates, and locations. The web interface built with SvelteKit provides a responsive timeline view, albums, shared albums with configurable permissions, public sharing links with optional passwords and expiry dates, partner sharing for family libraries, and a global map plotting photos by GPS coordinates. Administrative features include multi-user support with per-user storage quotas, OAuth integration, API key management, and a user-defined storage structure for organizing files on disk. The architecture uses PostgreSQL for metadata, Redis with BullMQ for background job queues handling thumbnail generation, video transcoding, and smart search indexing, and exposes over 400 REST API endpoints documented via OpenAPI with auto-generated SDKs for web, mobile, and CLI clients. 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.

Backed by Alibaba's AIDC team and carrying over 27,700 GitHub stars, Pixelle-Video turns a single text prompt into a publish-ready short video in approximately three minutes — handling scriptwriting, image generation, voice narration, music selection, subtitle overlay, and final MP4 export in one automated pipeline. The engine supports multiple LLM backends for script generation including GPT-4, Qwen, DeepSeek, and local Ollama deployments, while image and video creation routes through either self-hosted ComfyUI workflows, cloud-based RunningHub pipelines, or direct API connections to DashScope Wan, OpenAI, Seedream, Seedance, and Kling AI. Text-to-speech synthesis uses Edge-TTS, Index-TTS, and other mainstream engines with multi-language voice profiles. Five distinct pipelines cover Quick Create, Standard, Digital Human Avatar broadcasting, Image-to-Video transformation, and Motion Transfer from reference video. The Streamlit web UI on port 8501 provides a visual workflow builder with template selection across portrait (1080x1920), landscape (1920x1080), and square formats, while the FastAPI server on port 8000 exposes a REST API with endpoints for async video generation, task polling, content scripting, TTS and image generation, template listing, and health checks. History persistence tracks all completed generations. HTML-based visual templates support static, image-overlay, and AI-video styles with customizable prompt prefixes. The modular architecture lets operators swap any atomic capability — image model, video model, TTS engine, or VLM — by editing a workflow JSON file without touching Python code. 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.

Bifrost is an open-source AI gateway that unifies 23+ LLM providers into a single OpenAI-compatible endpoint with automatic failover, semantic caching, and built-in cost governance, so one provider going down never takes your production AI application with it. Point your existing OpenAI or Anthropic SDK at Bifrost's local endpoint and gain access to OpenAI, Anthropic, AWS Bedrock, Google Vertex, Azure, Groq, Mistral, and Ollama without changing application code. Define fallback chains that automatically switch providers when one returns errors or exceeds latency thresholds, keeping response times stable during outages. The built-in web dashboard at port 8080 lets you configure providers, create virtual API keys, monitor live request traffic, and review analytics without editing configuration files. Semantic caching combines exact hash matching with vector similarity search via Weaviate, serving cached responses for identical or paraphrased prompts in sub-millisecond time to cut costs on repetitive workloads. The MCP gateway connects AI agents to external tools like filesystems, databases, and web APIs, exposing them to clients such as Claude Desktop and Cursor with per-key allow-lists. Four-tier budget hierarchy at customer, team, virtual key, and provider levels enforces spend caps, rate limits, and model restrictions across your organization. Extend functionality through custom Go plugins for analytics, monitoring, or security middleware. Native Prometheus metrics and OpenTelemetry distributed tracing give operations teams full production observability. 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 AI application is burning through API tokens faster than you can refresh the billing page, and you have no idea which prompt template is responsible. OpenLIT plugs that visibility gap with a self-hosted observability platform built specifically for LLM workloads. Add one line of code to instrument 90+ LLM providers, agent frameworks, and vector databases, then watch every request flow through a tracing dashboard that shows tokens consumed, latency measured, and dollars spent per call, per model, per environment. The requests view lists every LLM interaction with provider, model, cost, and token breakdown in a filterable table, while the trace detail panel lets you drill into individual spans to read the exact prompt sent and response received. Prompt Hub turns prompts into versioned artifacts you deploy, rollback, and A/B test without touching application code. OpenGround compares models side by side on the same input, so you can evaluate cost-versus-quality tradeoffs before committing to a provider. Automated evaluations run LLM-as-a-judge scoring on live production traces, flagging hallucinations, bias, and toxicity in real time. The Vault stores and rotates API keys centrally so secrets stay out of your codebase. Custom dashboards let you build drag-and-drop monitoring views with charts, stat cards, and tables backed by SQL queries against ClickHouse. GPU utilization, memory, temperature, and power metrics feed into the same platform for end-to-end infrastructure visibility. 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.

Backed by Y Combinator and trusted by over 2,300 companies processing billions of observations monthly, Langfuse is the most widely adopted open-source platform for building, monitoring, evaluating, and debugging LLM applications. The hierarchical tracing engine captures every LLM call, tool invocation, retrieval step, and agent action as nested spans based on OpenTelemetry, with automatic cost calculation, latency tracking, and token usage attribution across sessions and users. Prompt Management separates prompts from code with versioned artifacts, label-based deployments, one-click rollbacks, and runtime SDK fetching with server-side caching, while linking every generation back to its exact prompt version for attribution analytics. The evaluation system supports LLM-as-a-judge scoring, heuristic code evaluators, user feedback collection, and manual annotation workflows that run automatically on production traces or against curated datasets. The Playground enables interactive prompt testing on real production inputs with side-by-side model comparison across providers. Datasets and Experiments define test cases for systematic benchmarking with comparative result visualization. Native SDKs for Python and TypeScript provide decorator-based instrumentation, while 100+ integrations cover LangChain, LlamaIndex, OpenAI SDK, LiteLLM, Vercel AI SDK, and any OpenTelemetry-instrumented framework. The analytics dashboard surfaces cost breakdowns, quality scores, latency percentiles, and usage trends across models and prompt versions. 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 over 77,000 GitHub stars, Crawl4AI is the most-starred open-source web crawler on GitHub and the go-to tool for converting the web into AI-ready data. Built on Python and Playwright, it transforms any website into clean Markdown with headings, tables, code blocks, and citation hints optimized for LLM ingestion, or extracts structured JSON via CSS selectors, XPath expressions, or direct LLM-based schema extraction through OpenAI, Anthropic, and Ollama providers. The self-hosted Docker server exposes a REST API on port 11235 with endpoints for crawling, streaming results, screenshots, PDF generation, JavaScript execution, and LLM-powered extraction. Version 0.9.x introduced secure-by-default operation with mandatory JWT authentication, strict request validation, declarative hooks replacing inline code, and bounded job queues. Adaptive crawling uses information foraging algorithms to determine when sufficient data has been gathered, while deep crawl mode traverses link graphs intelligently. The async browser pool manages concurrent sessions with stealth plugins, proxy rotation, custom headers, and session persistence for authenticated scraping. A built-in MCP server enables direct integration with Claude, ChatGPT, and Cursor for AI-driven web research workflows. Content filtering applies BM25 and TF-IDF relevance scoring to extract only pertinent sections from noisy pages. 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.

With over 29,000 GitHub stars and deep integrations into LangChain, LlamaIndex, and CrewAI, Chroma has become the default vector database for developers building retrieval-augmented generation pipelines and AI agent memory systems. Its core API consists of just four functions — create, add, query, and delete — making it the fastest path from zero to semantic search, while the underlying Rust engine handles tokenization, embedding, HNSW indexing, and similarity scoring automatically. Chroma supports dense vector search via HNSW with configurable distance metrics including L2, cosine similarity, and inner product, sparse vector search using SPLADE, full-text BM25 keyword search, and regex matching, all combinable in hybrid queries through a single unified interface. Metadata filtering at query time uses MongoDB-style operators including $eq, $ne, $gt, $lt, $in, and logical combinators $and and $or, enabling precise result scoping without post-processing. The multimodal pipeline powered by OpenCLIP embeds text and images into a shared vector space, allowing cross-modal retrieval where text queries return relevant images and vice versa. Deployment options range from embedded mode via PersistentClient for notebooks and prototypes, to client-server mode with Docker for production, to Chroma Cloud for serverless scalability. Official Python and JavaScript SDKs provide identical APIs, and embedding function integrations support OpenAI, Cohere, Hugging Face, Google, Ollama, and custom models. 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.

Images, text, audio, video, HTML, PDFs, and time series, labeled in one tool with a standardized output format: Label Studio is the open-source data labeling platform for building training datasets. Computer vision tasks cover classification, object detection (boxes, polygons, ellipses, keypoints), and semantic segmentation; audio work spans transcription, speaker diarization, and emotion recognition; NLP handles named entity recognition and document classification with taxonomies up to 10,000 classes; and GenAI workflows support LLM fine-tuning data and RLHF response ranking. Labeling interfaces are fully configurable with an XML-like templating language, so the UI matches the task instead of the reverse. The ML backend SDK turns any model into a connected web server for pre-annotation (model predicts, humans verify), interactive labeling (real-time predictions as annotators draw regions or highlight text), and model evaluation - cutting annotation time dramatically on large datasets. Data imports from S3, GCS, or file uploads; the Data Manager filters and explores tasks; exports convert to the format your ML library expects via label-studio-converter. Multi-user accounts tie every annotation to its author, and webhooks, a Python SDK, and REST API embed labeling into any pipeline. Self-hosting keeps proprietary training data - often a company's most sensitive asset - entirely on your infrastructure.

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.

OpenLLM serves any large language model as an OpenAI-compatible API endpoint from a single CLI command, handling model download, backend selection, quantization, and port binding automatically. It supports the full spectrum of popular models including Llama 3.3, Qwen2.5, DeepSeek, Mistral, and Phi3, choosing between vLLM and PyTorch inference backends based on hardware capabilities. When vLLM is available, continuous batching with PagedAttention achieves up to 23x throughput improvement over naive serving, while GPTQ and bitsandbytes quantization reduces memory requirements for GPU-constrained deployments. The server exposes a RESTful API on port 3000 with full OpenAI client library compatibility, enabling drop-in replacement for commercial providers in any application using the standard chat completions format. A built-in web chat UI at the /chat endpoint provides immediate interactive testing without external clients. Custom model repositories allow teams to maintain private catalogs of fine-tuned models alongside the default repository that tracks the latest releases. Deployment workflows generate production-ready Docker images automatically, with Kubernetes manifest support for orchestrated scaling. Native integration with LangChain and LlamaIndex supports RAG pipelines, Transformers Agents enables tool-calling workflows, and HuggingFace Hub handles model discovery. Server-Sent Events enable real-time token streaming across all API endpoints. Backed by BentoML's production ML infrastructure. 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.

Cognee gives AI agents persistent long-term memory that survives across sessions, replacing the traditional stack of separate graph, vector, and session databases with a unified engine running on a single PostgreSQL instance. The memory-native API exposes four verbs (remember, recall, forget, and improve) enabling agents to persist context, retrieve cited answers, prune outdated knowledge, and self-improve from feedback. Under the hood, Cognee combines pgvector embeddings with a PostgreSQL-native graph store and cognitive-science-grounded ontology generation, delivering hybrid retrieval that fuses semantic similarity, structural graph traversal, and lexical search in a single query. Integrations span Claude Code, Cursor, LangGraph, OpenAI Agents, and any MCP-compatible client through a dedicated MCP server on port 8001, while the Python and TypeScript SDKs provide direct programmatic access. The platform supports swappable backends including Neo4j, FalkorDB, Qdrant, ChromaDB, Weaviate, Milvus, and LanceDB for teams with existing infrastructure. Built-in OpenTelemetry tracing, an experimental dashboard with knowledge graph visualization, multi-tenant user isolation, and audit trails ensure production readiness. Deploy via Docker Compose with optional profiles for PostgreSQL, Neo4j, Redis, and the web frontend. Reached v1.0 in April 2026 with 30,000+ stars. 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.

Backed by Y Combinator (S24) and processing traces from thousands of AI agents in production, Laminar is the open-source observability platform that treats agent debugging as a first-class engineering discipline rather than an afterthought. Its OpenTelemetry-native SDK auto-instruments Vercel AI SDK, LangChain, OpenAI, Anthropic, Gemini, Browser Use, Stagehand, Mastra, Pydantic AI, and the OpenAI Agents SDK with a single line of code, capturing every LLM turn, tool call, and sub-agent delegation as nested spans with full input/output data and token costs. The Signals engine lets you describe failures in plain language — "agent is stuck in a loop" or "tool returned empty results" — then reads every trace and alerts via Slack when it detects a match. A built-in debugger records runs and replays them from cache so each iteration takes seconds, designed for Claude Code, Cursor, or Codex to drive the repair loop via the MCP server or CLI. Run code-first evaluations in Python or TypeScript locally or in CI/CD pipelines, build datasets from production traces, and query everything with raw SQL through custom dashboards, the in-app editor, or your coding agent. The Rust backend delivers 20x trace compression, a custom real-time streaming engine, ultra-fast full-text search, and gRPC ingestion, while ClickHouse powers columnar analytics and PostgreSQL stores application state. 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.