skills.rest
baleen37
baleen37COMMUNITY
Explore all skills by this author →
GitHub

autoresearch

Automate autonomous experiment loops for optimization tasks with resumeable JSONL state.

4|Updated Jan 9, 2026
Search Tags:#automation#logging#optimization#experimentation#benchmarks#autoresearch
One-click install
npx skills add https://github.com/baleen37/bstack --skill autoresearch-baleen37
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/baleen37/bstack/tree/main/plugins/autoresearch/skills/autoresearch
Command: npx skills add https://github.com/baleen37/bstack --skill autoresearch-baleen37

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Autoresearch automates the design, execution, and logging of autonomous optimization experiments, enabling continuous improvement without manual intervention.

Core Features & Use Cases

  • Setup and manage autonomous experiment loops that progressively improve a target metric.
  • Read source code, log results to a structured state file, and resume sessions after interruptions.
  • Use for optimization, benchmarking, and iterative experimentation across software projects.

Quick Start

Start an autonomous optimization loop for a given goal.

Frequently Asked Questions about autoresearch

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

FAQPage Schema
How do I automate optimization experiments to run continuously without manual intervention?▼

You can automate optimization experiments by setting up autonomous experiment loops that progressively improve a target metric, log results to structured JSONL state, and resume sessions after interruptions without manual intervention.

Can I resume an optimization benchmark session after an interruption?▼

Yes, you can resume optimization benchmark sessions after interruptions. The system maintains JSONL state in a .autoresearch directory, allowing iterative experimentation to continue from the last logged result.

What is autonomous performance tuning for software projects and how does it work?▼

Autonomous performance tuning automates the design, execution, and logging of iterative optimization experiments across software projects, using goal-driven benchmarks to progressively improve performance metrics without manual oversight.

Do I need any specific dependencies to run autonomous experimentation loops?▼

No external dependencies are required to run autonomous experimentation loops. The setup enforces YAML frontmatter with name and description in SKILL.md and uses a run.sh script to manage execution.

What's the best way to log structured results from iterative benchmarking experiments?▼

The best way to log structured results from iterative benchmarking is writing them to a JSONL state file, which supports resumeable runs and enables tracking of performance tuning progress across software projects.

Are there limitations to using autonomous experiment loops for goal-driven benchmarks?▼

Autonomous experiment loops for goal-driven benchmarks are limited to optimization tasks that can be measured iteratively, requiring a defined target metric and source code that supports progressive performance tuning.

Related Skills

Explore More Data & Analytics Skills→
Thanh-aperoThanh-apero
1

ck:autoresearch

Iteratively adjust code or configuration to improve measurable metrics with Git rollback.

Community
Advanced
revfactoryrevfactory
68

worklog

Summarize Claude Code sessions into weekly Markdown worklogs grouped by project.

Community
Intermediate
lengoctuong2005lengoctuong2005
5

weights-and-biases

Automate logging and tracking of machine learning experiments with wandb.

Community
Advanced
0xHoneyJar0xHoneyJar
2

run-mode

Execute autonomous sprint loops with PR lifecycle orchestration and safety guardrails.

Official
Advanced