Lights-On Software Factory Boilerplate for Claude Code

A production-tested boilerplate for building autonomous software factories with Claude Code launched on GitHub today, offering developers an 80% reduction in token costs compared to unoptimized workflows. The open-source framework, created by Gojko Adzic, packages the orchestration architecture developed for Votito into a reusable starting point for teams looking to automate development pipelines.

Architecture and Cost Optimization

The boilerplate introduces an orchestration layer between development tasks and Claude Code API calls. This intermediary component analyzes work requests, batches related operations, and eliminates redundant context passing that inflates token usage in naive implementations. According to the project documentation, the Votito deployment achieved approximately 80% token cost reduction by preventing duplicate code analysis and minimizing context window bloat across sequential operations.

The framework handles common automation patterns including code generation, testing, refactoring, and deployment preparation. By maintaining state between operations and reusing analysis results, the orchestrator avoids repeatedly sending identical code contexts to Claude Code for similar tasks.

Release Date and Availability

Officially released on October 20, 2026, the boilerplate is available under an open-source license on GitHub. The repository includes configuration templates, orchestration logic, and integration examples that teams can adapt to their specific development workflows. Documentation covers setup procedures for various CI/CD environments and provides guidance on customizing the orchestration rules for different project types.

Lights-On Factory Concept

The term "lights-on" refers to software factories that operate continuously with minimal human oversight, contrasting with traditional "lights-out" factories that aim for zero human involvement. This approach acknowledges that development automation benefits from strategic human checkpoints while automating repetitive analysis and code manipulation tasks.

Key capabilities include:

• Automated code review and refactoring suggestions
• Test generation and coverage analysis
• Documentation updates synchronized with code changes
• Dependency management and security scanning

Production Validation

The architecture draws directly from Votito's operational deployment, where the orchestrated approach proved viable for real-world development cycles. The production case study demonstrates that token efficiency gains translate to meaningful cost savings at scale, particularly for teams running continuous integration workflows with frequent Claude Code invocations.

The boilerplate preserves the core orchestration patterns while removing Votito-specific customizations, creating a foundation that development teams can extend based on their toolchains and processes.

What This Means

This release lowers the barrier for teams exploring autonomous development automation by providing a proven architectural pattern. The 80% token cost reduction addresses one of the primary concerns preventing wider adoption of AI-assisted development workflows. Organizations can now experiment with lights-on factories using battle-tested orchestration logic rather than building optimization layers from scratch. As Claude Code and similar AI coding assistants mature, efficient orchestration frameworks will likely become standard infrastructure for development teams managing automation at scale.