autoresearch

Name: autoresearch
Availability: InStock
Author: Conversion

Run baseline-scored iteration loops with keep or reject outcomes on editable artifacts.

49|7|Updated May 4, 2026

Search Tags:#benchmarking #experiment design #quality gates #skill-eval #autoresearch #ablation testing

Ready to run

npx skills add https://github.com/agencia-conversion/agentic-seo-skills --skill autoresearch-agencia-conversion

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/agencia-conversion/agentic-seo-skills/tree/main/skills/autoresearch
Command: npx skills add https://github.com/agencia-conversion/agentic-seo-skills --skill autoresearch-agencia-conversion

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Autoresearch helps you improve a single, editable surface with a controlled iteration loop using baseline-scored metrics and an explicit keep or reject decision.

Core Features & Use Cases

Baseline-first evaluation: scores the starting artifact before proposing any variation.
One-change-per-iteration experimentation: changes only the declared editable surface per iteration to isolate cause and effect.
Committed metrics and gates: locks metrics up front, prevents lowering decision/check gates, and stops correctly using threshold/plateau/max-iter rules.
Evidence discipline: separates raw evidence from synthesis and keeps reviewable run notes under the skill-evals directory.

Quick Start

Ask for an experiment run to improve a specific skill file, and instruct the agent to use skill-eval mode with that file as the only editable surface and to save notes under .context/skill-evals/autoresearch/<run-id>/.

Frequently Asked Questions about autoresearch

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

FAQPage Schema

How do I run controlled keep or reject experiments for prompt tuning?▼

Controlled keep or reject experiments require a baseline score for your starting artifact, a single editable surface per iteration to isolate cause and effect, and committed metrics with explicit threshold, plateau, or max-iteration stop rules.

What is baseline-first scoring in experiment design?▼

Baseline-first scoring in experiment design is the process of evaluating and recording the performance of your original artifact before proposing any variations. This locked baseline provides the control metric needed to measure whether a single iterative change should be kept or rejected.

How do I isolate cause and effect during ablation testing on content structures?▼

To isolate cause and effect during ablation testing, restrict modifications to one declared editable surface per iteration loop. This single-change discipline ensures score variations are directly attributable to that specific modification.

Can I use autoresearch for benchmarking agentic SEO skills?▼

Yes, you can use autoresearch for benchmarking agentic SEO skills, prompts, and content structures. It applies a controlled iteration loop with committed metrics and quality gates to evaluate and improve a single editable skill file per run.

When should I not use a single-change-per-iteration experimentation loop?▼

You should not use a single-change-per-iteration experimentation loop if you need to modify multiple surfaces simultaneously or if you lack a defined editable artifact. The method requires one committed editable surface and locked metrics up front to function correctly.