eval-authoring

Generate and curate eval scenarios to validate Tessl skills.

Updated Apr 17, 2026
One-click install
npx skills add https://github.com/jbaruch/coding-policy --skill eval-authoring
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eval-authoring
Source: https://github.com/jbaruch/coding-policy/tree/main/skills/eval-authoring
Command: npx skills add https://github.com/jbaruch/coding-policy --skill eval-authoring

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the creation and quality assurance of eval scenarios to assess and improve Tessl skills.

Core Features & Use Cases

  • Scenario generation: produce diverse eval scenarios aligned with the REVIEW_CHECKLIST.
  • Review and quality checks: evaluate tasks and criteria for bleeding, leaking, and coverage gaps.
  • Coverage analysis: identify gaps and guide focused scenario expansion.
  • Score-driven iteration: drive improvements via lift analysis and targeted fixes.
  • Use Case: You are building a new skill and want to develop robust evals, review existing evals, or expand coverage.

Quick Start

Run tessl scenario generate to begin creating eval scenarios, then review and refine each scenario per the checklist.

Frequently Asked Questions about eval-authoring

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate eval scenarios to validate software skills?

Generate eval scenarios by producing diverse test cases aligned with a review checklist, then apply quality checks to assess tasks for bleeding, leaking, and coverage gaps to validate skills.

What are bleeding and leaking in eval scenario quality checks?

Bleeding and leaking are quality check criteria evaluated against a review checklist to identify scenario misalignment or data exposure, ensuring generated eval scenarios accurately validate intended skill behaviors.

How do I analyze eval coverage gaps and iterate on test scenarios?

Analyze eval coverage gaps by reviewing generated scenarios against checklist criteria, then use score-driven lift analysis to guide targeted fixes, iteratively expanding scenario coverage and improving validation quality.

Can I use this approach to review existing eval scenarios for a new skill?

Yes, you can review existing eval scenarios by applying checklist-based quality checks to assess bleeding, leaking, and coverage gaps, evaluating whether current test cases sufficiently validate skill behaviors.

What is the best way to automate test case creation for skill development workflows?

Automate test case creation by running a scenario generator to produce diverse eval scenarios aligned with a review checklist, then apply quality checks and feedback loops to ensure comprehensive coverage across development workflows.