post-eval

Verify eval batch results and generate a concise summary.

Updated Jul 2, 2026

Search Tags:#automation #evaluation #dashboard #pipeline #data-analysis #post-eval #results-verification

Ready to run

npx skills add https://github.com/SamyakJhaveri/loam --skill post-eval

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: post-eval
Source: https://github.com/SamyakJhaveri/loam/tree/main/seed/_research/skills/post-eval
Command: npx skills add https://github.com/SamyakJhaveri/loam --skill post-eval

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Post-batch analysis pipeline that runs after an eval batch completes. Use after /eval-run or /overnight-eval finishes. Verifies results, runs analysis scripts, refreshes dashboard, writes summary. Does NOT launch new eval runs.

Core Features & Use Cases

Verifies results from eval batches
Runs analysis scripts to classify and summarize outcomes
Refreshes dashboards and writes a concise summary for reporting
Use Case: Apply after an eval batch across models/configurations to ensure quality and visibility

Quick Start

Invoke the post-eval pipeline after an eval batch completes to verify results and generate a summary.

Frequently Asked Questions about post-eval

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

FAQPage Schema

How do I automate post-batch evaluation analysis and dashboard updates?▼

Automating post-batch evaluation analysis involves running a pipeline that verifies eval batch results, executes analysis scripts, refreshes dashboards, and generates a concise structured summary for reporting.

What is the best way to verify results after an overnight eval run?▼

Verifying results after an overnight eval run requires an automated post-eval pipeline that performs integrity checks, validates outputs across multiple models and configurations, and classifies outcomes to ensure quality.

Does the post-eval pipeline launch new evaluation runs?▼

No, the post-eval pipeline does not launch new evaluation runs. It strictly processes completed eval batches to verify results, run analysis scripts, refresh dashboards, and write structured summaries.

How do I generate a summary report after completing an evaluation batch across multiple models?▼

Generating a summary report after an evaluation batch uses a post-eval pipeline to run script-driven analysis, classify outcomes across models and configurations, and write a concise structured summary for visibility.

Can I use post-eval analysis to validate outputs and refresh dashboards for multiple configurations?▼

Yes, post-eval analysis applies across multiple models and configurations to validate outputs, execute integrity checks, refresh dashboards, and generate structured summaries ensuring quality and visibility.

What steps are included in a post-eval analysis pipeline?▼

A post-eval analysis pipeline implements sequential steps for integrity checks, result verification, script-driven analysis, dashboard updates, and structured summary generation to process completed eval batches.