What problem does it solve?
Measuring whether an agentic workflow actually achieves its intended repository outcome is hard: raw activity counts and agent self-assessments do not prove attainment. This Skill designs a deterministic operational-value grader that scores each workflow run against frozen, evidence-based criteria.
Core Features & Use Cases
- Grader Generation: Creates one executable Bash evaluator at
.github/graders/WORKFLOW-NAME-operational-value.sh implementing the schema-version 4 definition contract with --definition, --metric, and --grade-run interfaces.
- Evidence & Baseline Freezing: Recovers adoption-time intent from the workflow's first commit, freezes accepted evidence, maturation rules, and an optional pre-adoption baseline for delta computation.
- Verification & Reporting: Validates evaluators with the bundled verification script, supports regrading historical runs, and builds JSON/SVG/Markdown reports across all completed runs.
- Use Case: After deploying an issue-triage agentic workflow, use this Skill to build a grader that scores each run by whether the assigned issue actually received the intended repository outcome, then track weekly attainment trends against a frozen baseline.
Quick Start
Ask the agent to design an operational-value grader for your repository and workflow by running /operational-value-designer with your OWNER/REPO and workflow name.
