About

Ruminate

Make the skills you rely on better with every real run.

Ruminate keeps selected dogfooding skills aligned between their working source and installed copy. Before a tracked skill is used or changed, it checks whether main has moved. A roughly ten-minute cooldown avoids repeated checks during a stretch of work, and each use records the exact source revision behind the behavior.

When real use exposes a rough edge, Ruminate turns it into an actionable improvement. It protects local edits, reconciles changes safely on main, checks for concurrent updates before committing, and refreshes the installed copy for the next run.

Built for deliberate iteration

• Track only the skills you explicitly choose to use and improve.
• Check before use or change, with no startup hooks, daemon, scheduled updater, or marketplace integration.
• Preserve uncommitted local work and never mistake a failed source check for a current version.
• Improve the working source on main, then install that revision for the next run.

Ruminate never updates skills outside your explicit list and does not promise to intercept every host action. It keeps the chosen skills in a focused loop: use, notice, improve, and use again.

Coming soon. Ruminate Pro is being developed. There is no supported public installation or purchase option yet.