refactor: streamline agentv-bench SKILL.md, fix grading bugs#848

Merged

christso merged 10 commits intomainfrom

refactor-bench-skill

Mar 29, 2026

Commits on Mar 29, 2026

refactor: streamline agentv-bench SKILL.md orchestration flow

christso
and
claude
committed
fix: address code review findings

christso
and
claude
committed
fix: remove internal `kind: "agent"` references from skill docs

christso
and
claude
committed
refactor: remove Python wrapper scripts from agentv-bench

christso
and
claude
committed
refactor: condense SKILL.md frontmatter description

christso
and
claude
committed
feat: add --grader-type flag to pipeline run

christso
and
claude
committed
docs: document --grader-type flag in skill references

christso
and
claude
committed
refactor: pipeline run defaults to skipping graders

christso
and
claude
committed
fix: use skill-trigger assertion in description optimization examples

christso
and
claude
committed
fix: correct skill-trigger assertion syntax

christso
and
claude
committed