autoresearch-create

Set up and run autonomous experiment loops with logging and version control.

Updated Feb 27, 2026

Search Tags:#logging #metrics #baseline #version-control #experiments #autoresearch

Ready to run

npx skills add https://github.com/cfcosta/duskpi --skill autoresearch-create-cfcosta

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: autoresearch-create
Source: https://github.com/cfcosta/duskpi/tree/main/skills/autoresearch-create
Command: npx skills add https://github.com/cfcosta/duskpi --skill autoresearch-create-cfcosta

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the setup and execution of autonomous experiments to identify effective optimizations without manual reconfiguration.

Core Features & Use Cases

Autonomous loop creation: configure, baseline, and run iterative experiments with minimal input.
Metric-driven evaluation: capture primary and secondary metrics and decide on keeping or discarding results.
Automatic logging & version control: records decisions, ensures reproducibility, and rolls back unsuccessful runs.

Quick Start

Provide goal, command, and metric details, then trigger autoresearch to start the autonomous loop immediately.

Frequently Asked Questions about autoresearch-create

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

FAQPage Schema

How do I automate iterative experiments to optimize a specific goal?▼

You can automate iterative experiments by configuring an autonomous loop that proposes, executes, and evaluates runs against baselines using metrics. This setup requires minimal input to continuously optimize your specified goal without manual reconfiguration.

How does metric-driven evaluation work for autonomous experiment loops?▼

Metric-driven evaluation captures primary and secondary metrics during each experiment run to decide whether to keep or discard results. This mechanism ensures that only effective optimizations are retained based on your defined performance criteria.

Can I automatically log and version-control experiment state during optimization runs?▼

Yes, autonomous experiment loops support automatic logging and version-control to record decisions and ensure reproducibility. This allows you to track experiment state and roll back unsuccessful runs effectively.

What do I need to set up an autonomous experiment loop with minimal configuration?▼

You need to provide your goal, command, and metric details to trigger the autonomous experiment loop. This initial input generates the necessary configuration and state management files to start iterative improvement immediately.

Does this autonomous experiment tool require any external dependencies to run?▼

No, this autonomous experiment tool operates without external dependencies. It independently manages the experiment state, logging, and version control required to run and evaluate your optimization tasks.

When should I use autonomous experiments instead of manual optimization?▼

You should use autonomous experiments for iterative improvement tasks where you need to identify effective optimizations without manual reconfiguration. It is ideal when you want to automatically evaluate results against baselines and discard unsuccessful runs.