Similar to the older BoxMaker, this ruby gem generates PDFs that can be used as a basis for cutting boxes on a typical laser cutter. The intention was to create an extensible, well tested, and modern ruby gem for generating PDF templates used in laser cutting.

Keeps a project's specifications, plans and pull requests joined up, and drives specialist agents over them in parallel. A feature's state is its folder name under .plans, so a transition renames a directory rather than updating a row, and nothing can claim a phase whose document is missing.

This compact ruby gem is purely CLI utility: it's meant to be used by a team of AI agentsexecuting along one or more the plans in a given repo. Using worktrees and parallelism it's easy to 10x the speed of develpoment of software development compared to even a single agent working on it. Plus each agent can specialize. Such agents require globsl shared locking mechanism to ensure they are not working on the same directory, or the same files in same worktree. That is exactly what this gem does. The CLI binary you invoke is called 'alock'which comes with sub-command 'completion' which you can load for BASH or ZSH. However, you are not very likely going toinvoke this gem directly. It's used by an Agentic Workflow gem 'agentilda' to protect shared resources. To try the entire system, it's recommended to download the repo https://github.com/kigster/agentilda-ai-setup which both installs the two gems, and offers a configuration file that installs a set of coding agents, skills, plugins, commands, from various guthub folders, or via running commands, and so on. In other words the repo's purpose is to ensure your agentic setup is identical from machine to machine, and by modifying the config file you can pick and choose your skills, your AGENT.md/CLAUDE.md file and so on.

Run multi-case evaluation suites against datasets using any model available through OpenRouter or the direct Jev adapter. Write an adapter that calls your application and a Scorer that measures the outcomes you care about. Compare models, or track how prompt and parameter changes affect results. Combine metrics such as accuracy, errors, latency, and cost into weighted scores, then review rankings and case-level results in an HTML report. Integrates with Braintrust Evals and several others. Paid integration is not required.

Keeps a project's specifications, plans and pull requests joined up, and drives specialist agents over them in parallel. A feature's state is its folder name under .plans, so a transition renames a directory rather than updating a row, and nothing can claim a phase whose document is missing.