Verified:GitHub

pdf

Evaluates pilot experiment outputs and generates a PASS/FAIL/INCONCLUSIVE verdict report.

Updated May 23, 2026

Search Tags:#report generation #experiment metrics #research wiki #pilot evaluation #verdict classification #Claude Code workflow #failure_reason

Ready to run

npx skills add https://github.com/duany049/skill-runtime-evolution --skill pdf-duany049

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pdf
Source: https://github.com/duany049/skill-runtime-evolution/tree/main/.claude/skills/exp-pilot-eval
Command: npx skills add https://github.com/duany049/skill-runtime-evolution --skill pdf-duany049

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes assets (resource) components.

What problem does it solve?

This Skill turns noisy pilot experiment outcomes into an actionable verdict, then keeps your research wiki synchronized so ideas move forward (or get eliminated) with consistent rules.

Core Features & Use Cases

Verdict evaluation: Reads pilot result JSONs and applies the Pilot Spec success criteria to classify outcomes as pass, fail, or inconclusive.
Wiki state update: Writes the derived pilot_result, sets failure_reason (with the mandatory [pilot] prefix) on failure, and transitions status to failed when appropriate.
Persistent reporting: Produces a terminal PILOT_VERDICT_REPORT and saves it to experiments/pilot/{slug}/report.md while appending wiki/log.md entries for traceability.

Quick Start

Run the pilot evaluation for an idea slug after a pilot run completes by invoking: /exp-pilot-eval <idea-slug> --auto.

Frequently Asked Questions about pdf

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

FAQPage Schema

How do I evaluate pilot experiment results and classify them as pass or fail?▼

To evaluate pilot experiment results, you compute metric summaries across pilot result seeds and compare them against the success criteria defined in the pilot spec to classify outcomes as pass, fail, or inconclusive.

How do I automate research wiki updates after a pilot run?▼

You can automate research wiki updates by enforcing forward-only lifecycle updates that write the derived pilot_result, set the failure_reason with a mandatory prefix, and transition the status field to failed when appropriate.

What is a pilot verdict report and how is it generated?▼

A pilot verdict report is a persistent record saved to the experiments directory that logs evaluation outcomes. It is generated by reading pilot logs and seeds, then appending traceability entries to the wiki log.

Can I use experiment metrics to enforce forward-only status transitions in an ideation workflow?▼

Yes, you can use experiment metrics to enforce forward-only status transitions in an ideation workflow by applying pilot spec success criteria and baseline comparisons to automatically transition wiki states.

What happens to the failure_reason field when a pilot evaluation yields a fail verdict?▼

When a pilot evaluation yields a fail verdict, the failure_reason field is set in the research wiki with a mandatory [pilot] prefix, and the idea status transitions to failed to maintain traceability.

What input data formats are required to classify pilot outcomes as inconclusive?▼

Classifying pilot outcomes as inconclusive requires reading pilot result JSONs, pilot logs, and the pilot spec to compute metric summaries and compare them against the defined success criterion and baselines.