The Ludopoly SDK operates through multiple access points — a command-line interface, a VS Code extension, a web-based IDE, and a REST API. The fastest path to a working package is through the CLI, which is what this guide covers. Whichever entry point you choose, the underlying production pipeline is identical: your intent is captured, refined into a specification, validated by the agent swarm, and compiled into a deployable package.

Install the Ludopoly CLI globally through npm. The package includes the orchestration client, authentication utilities, and local configuration tools.

Authenticate with your Ludopoly account to connect the CLI to the production pipeline. This step links your local environment to the platform's agent swarm and knowledge engine.

With authentication complete, you can request your first package. The generate command accepts a natural language description and an optional target chain flag.

The platform interprets your description, produces a requirements specification, and routes it through the eight-stage production pipeline. Within moments, you receive a complete package containing the smart contract, deployment scripts, test suites, and SDK bindings for your chosen platform.

Every generated package follows a consistent structure. The contract source lives alongside its compiled artefacts, a test suite that covers the primary execution paths, deployment configuration for the target chain, and typed SDK bindings for immediate integration into your application.

The security-report.md file documents the findings of the security agent — the same agent that holds veto authority over the production pipeline. If a critical vulnerability had been detected during generation, the package would not have been emitted at all.

When you submit a generation request, the platform performs a sequence of operations that would ordinarily require days of manual effort. The agent swarm decomposes your description into a formal specification. Specialised agents — one for contract logic, another for security, a third for gas optimisation — each produce candidate implementations. These candidates are evaluated through a weighted consensus mechanism, and only the output that meets all quality thresholds proceeds to the package factory for compilation and packaging.

The knowledge engine supports this process at every stage, supplying the agents with up-to-date information on EIP standards, known vulnerability patterns, gas cost benchmarks across chains, and best practices distilled from the broader blockchain ecosystem.