ar-archive

To archive completed autoresearch projects, move the project directory into autoresearch/archive while preserving its artifact bundle. This maintains long-term traceability by keeping files like results.tsv, last_eval.md, last_eval.json, and last_profile.md intact.

The best way to preserve autoresearch artifacts is to move the project into the archive directory. This action specifically retains key files like results.tsv and last_eval.md while pruning transient runtime artifacts to keep the archive clean and reproducible.

Yes, archiving an autoresearch project rewrites internal references in autoresearch.toml, program.md, and bootstrap.zsh. This ensures that the archived project's configuration remains consistent and correctly points to its new archived location.

Use an automated archive process when closing out or parking an autoresearch loop to ensure consistent archival records. It systematically moves the completed project and preserves its artifact bundle without manual file management.

Yes, the autoresearch archive process prunes transient runtime artifacts to keep archives clean and reproducible. It selectively preserves essential files like results.tsv, last_eval.md, last_eval.json, and last_profile.md while discarding temporary runtime data.