Score and gate skill improvements
os-eval-runnerskillsetup L3★3
richfrem/agent-plugins-skills ↗What it does
Run stateless skill evaluation iterations
Best for
Autonomously iterating skill improvements using empirical scoring gates instead of human feedback.
Inputs
- · phase-number
- · optimization-metric
Outputs
- · task-completion-status
Requires
- · Task/Subagent
- · git
- · Edit
- · Python
- · Bash
- · Read
Preconditions
- · Valid git repository initialized
- · Python 3.8+ available
- · Evaluation harness installed
Failure modes
- · Lab and master versions have diverged
Trust signals
- · Explicit verification gates
- · Multi-phase execution with gates
- · Empirical evaluation scoring