experiment-tracker

Consolidate validation experiments into a decision state with a recommended next experiment.

4|Updated Mar 6, 2026

Search Tags:#workflow #validation #experiment #signal-quality #decision-state #log-structured

Ready to run

npx skills add https://github.com/accolver/skill-maker --skill experiment-tracker-accolver

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: experiment-tracker
Source: https://github.com/accolver/skill-maker/tree/main/experiment-tracker
Command: npx skills add https://github.com/accolver/skill-maker --skill experiment-tracker-accolver

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

Messy validation histories often leave teams unsure what truly changed and what to test next. This skill synthesizes experiments, weighs signal quality, and produces a clear decision state plus a single recommended next experiment to drive forward.

Core Features & Use Cases

Summarizes experiments with labeled signals (strong, weak, vanity, contradictions).
Sets a concrete decision state (continue, narrow, pivot, or stop) and proposes one focused next experiment.
Outputs a readable summary and a structured log compatible with assets/experiment-log-template.json for reproducibility.
Useful when running multiple validation cycles, resolving conflicting evidence, or ensuring continuity across experiments.

Quick Start

Install or load this skill in an agent workflow and run a validation session to generate a decision-ready plan for the next experiment.

Frequently Asked Questions about experiment-tracker

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

FAQPage Schema

How do I consolidate conflicting validation experiments into a single decision state?▼

To consolidate conflicting validation experiments, you synthesize the test cycles by applying explicit signal labels like strong, weak, vanity, or contradictions. This process preserves continuity across noisy evidence and outputs exactly one recommended next experiment to drive forward.

What is the best way to track experiment history without rethinking validation from scratch?▼

Tracking experiment history without starting from scratch requires summarizing past tests with labeled signals and logging them in a structured format like assets/experiment-log-template.json. This establishes a concrete decision state to continue, narrow, pivot, or stop, ensuring reproducible continuity.

How do I evaluate signal quality across multiple noisy validation cycles?▼

Evaluating signal quality across noisy validation cycles involves categorizing evidence into explicit labels: strong, weak, vanity, or contradictions. Weighing these signals clarifies what truly changed and produces a focused, decision-ready plan for the next experiment.

Can I use a structured log to resolve conflicting evidence across product validation tests?▼

Yes, resolving conflicting evidence across product validation tests requires a structured JSON log compatible with assets/experiment-log-template.json. This format captures labeled signals and the resulting decision state, ensuring reproducibility and continuity across test cycles.

When should I stop, narrow, or pivot during a product validation workflow?▼

You should stop, narrow, pivot, or continue during a validation workflow after synthesizing all experiment signals into a single decision state. Weighing strong, weak, vanity, and contradictory evidence determines whether to proceed or change direction completely.

Why does my team keep getting conflicting evidence from different validation experiments?▼

Teams get conflicting evidence from validation experiments when noisy test results lack a consolidated decision state. Applying explicit signal labels and logging structured data prevents messy histories and ensures continuity rather than rethinking from scratch.