# Hubify Labs Docs

## Docs

- [Activity](https://hubify.com/docs/api/activity.md): Read the lab activity feed, agent actions, status changes, milestones, errors. The heartbeat of a running lab.
- [Agents](https://hubify.com/docs/api/agents.md): Configure, start, stop, and monitor AI agents in your lab's hierarchical multi-agent system.
- [Authentication](https://hubify.com/docs/api/authentication.md): API authentication, API keys, Clerk JWT tokens, scopes, and security best practices.
- [Backups](https://hubify.com/docs/api/backups.md): Trigger and list lab backups to Backblaze B2 or other destinations.
- [Chats](https://hubify.com/docs/api/chats.md): Promote a chat session to a knowledge entry, paper draft, or task.
- [Contributions](https://hubify.com/docs/api/contributions.md): List and create scientific contributions — discoveries that feed the novelty-review pipeline.
- [Costs](https://hubify.com/docs/api/costs.md): Query compute spend, approval queue, and cost breakdown by GPU type and experiment.
- [Datasets](https://hubify.com/docs/api/datasets.md): List and create research datasets linked to lab experiments.
- [Experiments](https://hubify.com/docs/api/experiments.md): Create, run, monitor, and manage GPU-powered experiments with automatic logging, QC gates, and status lifecycle.
- [Figures](https://hubify.com/docs/api/figures.md): Upload, manage, and organize figures, plots, charts, and diagrams linked to experiments and papers.
- [Files API](https://hubify.com/docs/api/files.md): Read, write, and delete lab files. Backs the local-agents and CLI workflows.
- [Knowledge Base](https://hubify.com/docs/api/knowledge.md): Create, search, and manage structured knowledge entries, the lab's growing wiki of concepts, parameters, sources, and comparisons.
- [Labs API](https://hubify.com/docs/api/labs.md): Create, read, update, and delete research labs via the REST API.
- [Memory](https://hubify.com/docs/api/memory.md): Read and write the four-layer memory system: messages, preferences, learnings, and decisions.
- [Messages](https://hubify.com/docs/api/messages.md): Send and receive chat messages, stream AI responses, and search conversation history.
- [API Overview](https://hubify.com/docs/api/overview.md): REST API overview, base URL, versioning, authentication, rate limits, and conventions.
- [Papers](https://hubify.com/docs/api/papers.md): List papers in a lab and run the 5-round publish loop (Mechanical QA, Cross-Model Review, Houston Method, Format, arXiv Prep).
- [Pods](https://hubify.com/docs/api/pods.md): Provision, monitor, and manage GPU compute pods, SSH access, cost tracking, and idle detection.
- [Projects](https://hubify.com/docs/api/projects.md): Manage research projects that group experiments, papers, and tasks.
- [Search](https://hubify.com/docs/api/search.md): Universal search across every entity in a lab, experiments, papers, agents, knowledge, surveys, tasks, figures, datasets, contributions.
- [Standups](https://hubify.com/docs/api/standups.md): Schedule and retrieve lab standups — structured agent progress check-ins.
- [Surveys](https://hubify.com/docs/api/surveys.md): List QC surveys — quality-control passes over experimental datasets.
- [Sync API](https://hubify.com/docs/api/sync.md): Manual GitHub sync trigger and webhook secret provisioning.
- [Tasks](https://hubify.com/docs/api/tasks.md): Create, assign, and manage tasks on the Kanban board, with status transitions, review workflows, priorities, and dependencies.
- [Telemetry](https://hubify.com/docs/api/telemetry.md): Track and query custom telemetry events from agents, CLI tools, and external runners.
- [Webhooks](https://hubify.com/docs/api/webhooks.md): Configure webhooks to receive real-time notifications when experiments complete, agents error, pods terminate, and more.
- [Changelog](https://hubify.com/docs/changelog.md): Version history for Hubify Labs, new features, improvements, and fixes.
- [hubify agent](https://hubify.com/docs/cli/agent.md): Agent management commands, add, configure, monitor, and manage your multi-agent research team.
- [hubify auth](https://hubify.com/docs/cli/auth.md): Authentication commands, login, logout, token management, and API key configuration.
- [hubify deploy](https://hubify.com/docs/cli/deploy.md): Deploy experiments to GPU pods, launch scripts, monitor progress, and check deployment status.
- [hubify experiment](https://hubify.com/docs/cli/experiment.md): Experiment lifecycle commands, create, run, monitor, and manage GPU-powered experiments.
- [Installation](https://hubify.com/docs/cli/install.md): Install the Hubify CLI on macOS, Linux, or Windows. Optionally configure the MCP server for Claude Code or Codex.
- [hubify lab](https://hubify.com/docs/cli/lab.md): Lab management commands, create, switch, list, update, and delete research labs.
- [hubify logs](https://hubify.com/docs/cli/logs.md): View and stream logs from experiments, agents, deployments, and system events.
- [CLI Overview](https://hubify.com/docs/cli/overview.md): The Hubify CLI gives terminal-first researchers full access to labs, experiments, agents, papers, and compute.
- [hubify paper](https://hubify.com/docs/cli/paper.md): Paper pipeline commands, create, draft, review, compile, and submit research papers.
- [hubify pod](https://hubify.com/docs/cli/pod.md): Manage GPU pods, create, list, SSH into, and stop compute instances for running experiments.
- [hubify review](https://hubify.com/docs/cli/review.md): Cross-agent peer review workflow, request reviews, view feedback, and approve or reject work.
- [hubify site](https://hubify.com/docs/cli/site.md): Manage your public lab site, publish, preview, and configure your research presence.
- [hubify standup](https://hubify.com/docs/cli/standup.md): All-hands agent standups, 3x/day roundtables with transcripts, action items, and escalation.
- [hubify task](https://hubify.com/docs/cli/task.md): Task queue commands, create, list, assign, prioritize, and complete tasks.
- [Core Concepts](https://hubify.com/docs/concepts.md): The building blocks of Hubify Labs: labs, agents, experiments, papers, and knowledge.
- [Agents](https://hubify.com/docs/concepts/agents.md): Hubify Labs uses a hierarchical multi-agent system with an orchestrator, lead agents, and workers, powered by cross-model peer review.
- [Compute](https://hubify.com/docs/concepts/compute.md): GPU compute in Hubify Labs, RunPod integration, pod management, cost optimization, and the GPU inference playbook.
- [Experiments](https://hubify.com/docs/concepts/experiments.md): Experiments are the atomic unit of research progress, GPU-powered tasks with automatic logging, QC gates, and reproducibility tracking.
- [Knowledge Base](https://hubify.com/docs/concepts/knowledge-base.md): Every lab has a Karpathy-style structured wiki that grows automatically as agents research, experiment, and discover.
- [Labs](https://hubify.com/docs/concepts/labs.md): Labs are isolated research environments, the top-level container for experiments, agents, papers, data, and compute.
- [Papers](https://hubify.com/docs/concepts/papers.md): The paper pipeline takes experiment results through drafting, peer review, LaTeX compilation, and arXiv submission.
- [FAQ](https://hubify.com/docs/faq.md): Frequently asked questions about Hubify Labs, pricing, capabilities, data, and research workflows.
- [Captain View](https://hubify.com/docs/features/captain-view.md): Your research command center, a unified dashboard showing experiments, agents, tasks, papers, and compute at a glance.
- [Data Explorer](https://hubify.com/docs/features/data-explorer.md): Interactive data exploration tool, sortable tables, column statistics, equation calculators, and node-tree visualization for experiment outputs.
- [Dynamic Surface Inference (DSI)](https://hubify.com/docs/features/dsi.md): Hubify's AI-native UI layer, agents render the Captain's view from a JSON spec, not a hardcoded layout. Coming soon.
- [Experiment Runner](https://hubify.com/docs/features/experiment-runner.md): Execute research experiments on GPU pods with automatic logging, checkpointing, QC gates, and reproducibility tracking.
- [GPU Compute](https://hubify.com/docs/features/gpu-compute.md): Manage GPU pods for experiments, provision H100/H200, monitor utilization, optimize costs, and SSH into running pods.
- [Knowledge Wiki](https://hubify.com/docs/features/knowledge-wiki.md): A structured, agent-maintained knowledge base that grows with your research, entities, concepts, sources, and comparisons.
- [Lab Site](https://hubify.com/docs/features/lab-site.md): Every lab gets a public research website at {slug}.hubify.app, automatically synced with your experiments, papers, and results.
- [Multi-Agent System](https://hubify.com/docs/features/multi-agent.md): A hierarchical agent team with cross-model peer review, reasoning-based routing, and full visibility into agent communication.
- [Novelty Scoring](https://hubify.com/docs/features/novelty-scoring.md): AI-powered assessment of how novel your findings are, from incremental to field-changing, calibrated against existing literature.
- [Orchestrator](https://hubify.com/docs/features/orchestrator.md): The AI orchestrator is your primary interface, a natural-language chat that routes tasks, manages agents, and keeps research on track.
- [Paper Pipeline](https://hubify.com/docs/features/paper-pipeline.md): End-to-end workflow from experiment results to arXiv-ready PDF, drafting, review, LaTeX compilation, and submission.
- [Agent Configuration](https://hubify.com/docs/guides/agent-configuration.md): Configure your multi-agent team, choose models, define specialties, set up cross-model review, and tune routing.
- [Create a Lab](https://hubify.com/docs/guides/create-lab.md): Step-by-step guide to creating your first research lab in Hubify Labs.
- [Cross-Lab Data Sharing](https://hubify.com/docs/guides/cross-lab-sharing.md): Enable the Lab Sovereignty Rule-compliant cross-lab read gateway. Share specific datasets and learnings with other labs without exposing private experiments.
- [Run Your First Experiment](https://hubify.com/docs/guides/first-experiment.md): Launch your first GPU-powered experiment in Hubify Labs, from setup to results in minutes.
- [GPU Setup](https://hubify.com/docs/guides/gpu-setup.md): Connect GPU compute providers, configure pod defaults, set budget limits, and optimize for cost.
- [Houston Method v2](https://hubify.com/docs/guides/houston-method-v2.md): The 9-step post-experiment protocol. Every experiment that completes triggers this ritual, pass or fail.
- [Lab Site Customization](https://hubify.com/docs/guides/lab-site-customization.md): Customize your lab's public research website, theme, pages, custom domain, and content management.
- [Local Agents (Claude Code, Codex, gh)](https://hubify.com/docs/guides/local-agents.md): Run any local AI coding agent against your Hubify lab. Edits sync both ways through GitHub.
- [Run the Publish-Ready Loop](https://hubify.com/docs/guides/publish-ready-loop.md): Trigger the autonomous 5-round publish loop on a paper. What each round checks, how to read the preflight scorecard, and how to export the arXiv package.
- [Publishing Papers](https://hubify.com/docs/guides/publishing-papers.md): End-to-end guide for taking experiment results through the paper pipeline to arXiv submission.
- [arXiv Integration](https://hubify.com/docs/integrations/arxiv.md): Submit papers to arXiv directly from Hubify Labs, packaging, validation, and submission.
- [Claude Code Integration](https://hubify.com/docs/integrations/claude-code.md): Use Hubify Labs directly from Claude Code CLI with the MCP server and built-in workflows.
- [Cursor Integration](https://hubify.com/docs/integrations/cursor.md): Use Hubify Labs from within Cursor IDE via the MCP server.
- [HuggingFace Integration](https://hubify.com/docs/integrations/huggingface.md): Push models, datasets, and results to HuggingFace Hub from Hubify Labs.
- [Modal Integration](https://hubify.com/docs/integrations/modal.md): Use Modal serverless GPUs with Hubify Labs for on-demand compute.
- [RunPod Integration](https://hubify.com/docs/integrations/runpod.md): Connect RunPod GPU pods to Hubify Labs for experiment execution.
- [Introduction](https://hubify.com/docs/introduction.md): Hubify Labs is the Scientific Discovery Platform, AI-powered research infrastructure for running experiments, writing papers, and publishing discoveries.
- [MCP Server Overview](https://hubify.com/docs/mcp/overview.md): The Hubify MCP server exposes labs, experiments, agents, and knowledge to AI coding assistants.
- [MCP Prompts](https://hubify.com/docs/mcp/prompts.md): Pre-built MCP prompts for research analysis, paper writing, experiment design, and review workflows.
- [MCP Resources](https://hubify.com/docs/mcp/resources.md): Available MCP resources for reading lab data, experiment results, and knowledge base entries.
- [MCP Setup](https://hubify.com/docs/mcp/setup.md): Install and configure the Hubify MCP server for Claude Code, Cursor, and any MCP host. Read and write lab data directly from your AI coding assistant.
- [MCP Tools Reference](https://hubify.com/docs/mcp/tools.md): All 48 Hubify MCP tools. Platform tools (unprefixed) drive lab data directly. Phase E ledger tools (hubify_* prefix) write to the agentEvents ledger and appear in the Hubify web app in real time.
- [Quickstart](https://hubify.com/docs/quickstart.md): Get your first research lab running in under 5 minutes.
- [Bring Your Own Agent](https://hubify.com/docs/skills/bring-your-own-agent.md): Hubify works with any agent that speaks MCP, Claude Code, Codex, Cursor, Cline, Continue, Zed, Aider. Same skills, same data, your tool of choice.
- [Skills Overview](https://hubify.com/docs/skills/overview.md): Hubify Skills are bring-your-own-agent recipes that turn Claude Code, Codex, or Cursor into a Hubify-aware research IDE, without ever installing the dedicated Hubify CLI.
- [Troubleshooting](https://hubify.com/docs/troubleshooting.md): Common issues and fixes for Hubify Labs, CLI, agents, pods, papers, and deployments.

## OpenAPI Specs

- [openapi](https://hubify.com/docs/openapi.yaml)

## Optional

- [GitHub](https://github.com/houstongolden/hubify)
- [Community](https://discord.gg/hubify)