create-new-experiment

Scaffolding an autoresearch experiment initializes a standardized folder structure with templates, git branch naming conventions, and registry updates. This ensures consistent experiment setup and discoverability across projects and teams.

The experiment template includes the src/experiments directory structure, program.md, run_accuracy.py templates, and SKILL.md frontmatter. It also creates optional scripts, references, and assets subdirectories depending on project needs.

No dependencies are required to bootstrap a new experiment setup. The scaffold operates independently to generate the directory structure, templates, and registry updates without external packages or components.

The experiment registry updates automatically during scaffolding to track new experiments. This provides a centralized record of initialized projects, enabling discoverability and consistency across different environments and team members.

Yes, the experiment scaffold supports optional scripts, references, and assets subdirectories. These are created alongside required directories like src/experiments and templates like program.md to accommodate specific project requirements.

The best way to standardize setup is using an automated scaffold that creates consistent directory structures, branch naming conventions, and validation workflows. This ensures all experiments follow the same initialization pattern across projects.