SkillHone Synthesis

This skill builds a generic closed-form Q/A dataset from real tool use.

Closed-form means each question has one intended answer, and the answer can be checked by code: exact match, regex, numeric tolerance, enum, or JSON field checks. The final benchmark should not require a human or an LLM judge for basic grading.

The central idea is simple:

1. Explore the target environment with tools.
2. Save what was learned as a graph of entities, relations, and reasoning walks.
3. Mine several Q/A samples from that graph.
4. Validate that each sample is answerable, unique, stable, hard enough, and mechanically gradable.

The skill is domain-agnostic. Concrete task files, tool wrappers, and domain examples live outside this skill.

Task Spec Coverage