eval-report

Aggregate session metrics and eval data into markdown or JSON reports.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/ItsMattG/property-tracker --skill eval-report
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: eval-report
Source: https://github.com/ItsMattG/property-tracker/tree/main/.claude/skills/eval-report
Command: npx skills add https://github.com/ItsMattG/property-tracker --skill eval-report

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Quickly turning scattered session metrics into a coherent evaluation report is manual and error-prone. This skill automates collection and synthesis of session metrics, checkpoint evals, and pass@k data to produce a clear performance picture.

Core Features & Use Cases

  • Metrics aggregation: reads session metrics, checkpoint evals, and pass@k data to compute key indicators.
  • Multi-source synthesis: combines metrics.jsonl, eval outputs, and pass@k tracker to produce a unified report.
  • Use Case: teams retrospective on model performance across sessions and branches, identifying strengths and gaps.

Quick Start

Run evaluate-report with your latest metrics.jsonl to generate a 7-day report.

Frequently Asked Questions about eval-report

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

FAQPage Schema
How do I automate generating an evaluation report from session metrics?

Aggregate session metrics by reading and combining metrics.jsonl, checkpoint eval outputs, and pass@k tracker data to compute totals, averages, pass rates, and trend insights. The skill synthesizes these sources to produce a unified markdown or JSON evaluation report.

What is the best way to combine pass@k data with checkpoint eval outputs?

Combine pass@k data with checkpoint evals by reading and synthesizing these multiple metric sources into a single performance picture. The skill calculates totals, averages, pass rates, and trend insights, outputting a comprehensive markdown or JSON report.

Can I generate a dashboard from metrics.jsonl for AI project retrospectives?

Generate evaluation dashboards from metrics.jsonl for AI project retrospectives by calculating key performance indicators across sessions and branches. The skill identifies model strengths and gaps by aggregating session metrics into clear reporting outputs.

Does the report generation process work without external dependencies?

The report generation process works without external dependencies, relying solely on internal logic to parse metrics.jsonl and eval JSONs. It processes pass@k tracker data independently to compute averages, pass rates, and trend insights directly.

How do I compute pass rates and trend insights from scattered session metrics?

Compute pass rates and trend insights by reading and combining scattered session metrics, checkpoint evals, and pass@k tracker data. The skill automates this collection to calculate averages and generate a unified markdown or JSON performance report.