What problem does it solve?
This Skill identifies where your configured fork fleet systematically differs from upstream defaults, so you can learn from collective operator choices and reduce configuration drift.
Core Features & Use Cases
- Weekly divergence digest across forks: Scans active forks of a target repo and compares each fork’s
aeon.yml against the running instance’s upstream skill defaults.
- Default-change recommendations (aggregate): Flags skills likely worth flipping upstream based on how often forks diverge (enable upward / disable downward), plus consensus for model/var/schedule.
- Actionable fleet signals: Produces a verdict, tiered bucket results (flip candidates, consensus, hotspots, emerging watchlist), and a list of heaviest customizers and fork-only experiments.
- Week-over-week change tracking: Persists a structured snapshot in
memory/topics/fork-skill-digest-state.json and computes deltas like new flips and new fork-only skills.
Quick Start
Run the skill with an optional target repo override by setting var to a GitHub repo like "owner/aeon" to generate the weekly divergence digest for that fork fleet.