KDCube Documentation

Start with what KDCube is, how environments and bundles work, or jump straight to quick start if you want to run it now.

KDCube is a platform and SDK for packaging AI applications into deployable bundles and running them in isolated environments. Bundles can mix plain Python, ReAct, Claude Code, LangGraph, CrewAI flows, ISO runtime execution, MCP, and live streaming.

Reveal documentation MCP endpoint

KDCube documentation is also available as a public, read-only MCP server on dev.kdcube.tech for MCP-capable clients.

https://dev.kdcube.tech/api/integrations/bundles/demo/demo-march/kdcube.copilot@2026-04-03-19-05/public/mcp/kdcube-doc

What KDCube Is

Environment model, bundles, Ingress, Proc, and the system overview.

Quick Start

Install the CLI, configure your first bundle, and run the platform locally.

Platform Architecture

Services, modules, request lifecycle, and multi-tenant execution model.

Communication System

SSE streaming, Socket.IO, REST API, and channeled output protocol.

Client Integration

Browser transport contract, SSE events, Socket.IO, authentication, and reconnect behavior.

SDK Integrations

Reusable Email and Telegram building blocks for bundles: account access, webhooks, Mini Apps, streaming, and signed downloads.

Configuration

Environment variables, bundle config, secrets injection, and runtime settings.

Deployment & Monitoring

Docker Compose, ECS Fargate, Kubernetes, and observability setup.

Platform Economics

Per-tenant cost accounting, budget caps, rate limiting, and usage tracking.

Security & Governance

Prompt injection mitigation, tool sandboxing, tenant isolation, and audit trails.

Build Bundles

Tier 1 bundle docs, ISO runtime, ReAct agent, MCP, streaming, configuration, storage, deployment, hot reload, and live feature gating.

ReAct V3 Agent

Timeline-native agent loop with tools, skills, compaction, pruning, memory recovery, sources pool, and ANNOUNCE.

New in stack

Claude Code Agent Support

Bundle-native Claude Code runtime with per-turn accounting, workspace management, git-backed sessions, safety model, and integration patterns.

Glossary

Canonical definitions of KDCube platform terms: ReAct loop, timeline, bundle, source pool, tenant boundary, supervisor, executor, and more.