autostar

Automate structured experimentation to optimize measurable artifacts across code, prompts, docs, and configurations.

39|2|Updated Mar 27, 2026

Search Tags:#automation #verification #memory #experimentation #inference #optimisation #autostar

Ready to run

npx skills add https://github.com/chrisvoncsefalvay/autostar --skill autostar

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: autostar
Source: https://github.com/chrisvoncsefalvay/autostar/tree/main/autostar-skill
Command: npx skills add https://github.com/chrisvoncsefalvay/autostar --skill autostar

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

a* provides a self-contained framework to turn vague improvement goals into a structured, measurable, and repeatable optimisation loop. It converts fuzzy quality aspirations into independent tracks, verifiers, and a disciplined mutation-evaluate-reflect cycle that learns from every attempt.

Core Features & Use Cases

Onboarding with explicit goal decomposition, track assignment, constraints, and budget before any work begins.
Multi-track rubric system supporting deterministic, external-tool, llm_judge, hybrid, and human_gate verifiers to quantify progress.
Durable memory and run artifacts that capture hypotheses, step results, and reflections to enable cross-run learning and improvement.

Quick Start

Describe your goal and I will set up tracks, verifiers, and an initial mission to start the first optimisation loop.

Frequently Asked Questions about autostar

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

FAQPage Schema

How do I automate structured experimentation for code and prompt optimisation?▼

Structured experimentation automates optimisation by decomposing goals into tracks and verifiers, then running a mutation-evaluate-reflect cycle. It applies to code, prompts, docs, and configurations to drive quantified improvements.

What types of verifiers can I use to measure optimisation progress?▼

Verifiers for measuring optimisation progress include deterministic, external-tool, llm_judge, hybrid, and human_gate types. These multi-track rubric systems quantify progress across independent tracks before any work begins.

How do I set up an iterative optimisation loop with measurable goals?▼

Setting up an iterative optimisation loop requires an onboarding phase to define goals, assign tracks, set constraints, and allocate a budget. The system then creates an initial mission to start the first mutation-evaluate-reflect cycle.

Can I use memory and run artifacts to enable cross-run learning during optimisation?▼

Memory and run artifacts capture hypotheses, step results, and reflections during optimisation. This durable memory enables cross-run learning by retaining context from every attempted improvement across multiple iterations.

What is the best way to turn vague improvement goals into a repeatable optimisation process?▼

The best way to turn vague improvement goals into a repeatable process is using a framework that enforces explicit goal decomposition and strict verifier types. It converts fuzzy quality aspirations into independent tracks within a disciplined loop.

Does AITune delegation work for inference tuning in optimisation tracks?▼

AITune delegation is an optional feature for inference tuning within optimisation tracks. It supports the multi-dimensional track system alongside other verifier types to drive quantified improvements during the experimentation cycle.