autoresearch

Automate codebase improvement loops with constraint discovery and baseline capture.

Updated Mar 24, 2026

Search Tags:#automation #git #plugin #iterative #code-improvement #llm-eval #autoresearch

Ready to run

npx skills add https://github.com/dhanesh/autoresearch --skill autoresearch-dhanesh

💬 Prompt for ChatGPT / Claude Chat▼

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill orchestrates an autonomous loop that discovers evaluation constraints, runs improvements, and produces a formal learning report to drive measurable codebase quality gains.

Core Features & Use Cases

Hybrid constraint discovery that analyzes your project and suggests concrete evaluation commands
Iterative improve-evaluate-iterate loop with baseline capture, progression tracking, and diminishing-returns detection
Rich, markdown reporting including token accounting, trajectory analysis, and final learning report
Git-based isolation on a dedicated autoresearch branch with pre-flight permission manifest and safe rollback

Quick Start

Start autoresearch to begin constraint discovery and empower your project to improve over successive iterations.

Frequently Asked Questions about autoresearch

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

FAQPage Schema

How do I automate iterative codebase improvement in git?▼

Autoresearch automates iterative codebase improvement by running an autonomous loop that captures baselines, applies enhancements, and tracks progression while isolating changes on a dedicated git branch with safe rollback capabilities.

How does constraint discovery work for code evaluation?▼

Hybrid constraint discovery analyzes your project to suggest concrete evaluation commands, establishing measurable criteria for the improve-evaluate-iterate loop to ensure codebase enhancements are quantifiable.

What is autonomous codebase optimization and when do I need it?▼

Autonomous codebase optimization orchestrates an automated loop for measurable quality gains across multiple axes. You need it when targeting structured constraints for iterative enhancements and formal learning reports in software projects.

Can I run automated code improvements without risking my main branch?▼

Yes, automated code improvements are enforced via git-based isolation on a dedicated branch using a pre-flight permission manifest, ensuring safety and reproducibility with comprehensive rollback options.

Does the iterative code enhancement loop detect diminishing returns?▼

Yes, the iterative code enhancement loop detects diminishing returns during progression tracking, halting unnecessary iterations and generating a rich markdown report with token accounting and trajectory analysis.

What are the limitations of using an autonomous loop for code quality gains?▼

Autonomous loops require a preflight permission manifest to operate safely and rely on discovered evaluation constraints. Without clear baseline capture or suitable constraints, the iterative enhancement process may lack measurable targets.