arbor
k-dense-ai1.1k installs
Improve research code with evidence from trials.
What it does
Improves an existing research artifact through isolated experiments, persistent hypothesis tracking, and held-out evaluation.
When to use it
Use for repeated improvement of a modifiable artifact with a measurable objective and an evaluator, rather than a single fix or unscored brainstorming.
How to use it
Give it an existing artifact and an improvement objective; it returns a runnable best candidate on a named branch and an evidence-backed report.
What you provide
- An existing project
Uses
Good to know
- starts one helper agent per item, so the count depends on the work
- advertises something the task does not need
Latest review
What you need · 1
Requires Python 3.10+ to run the bundled state manager.
Access · 1
This skill
arXiv
Read
Looks up the library's paper