skills.rest
alirezarezvani
alirezarezvaniCOMMUNITY
Explore all skills by this author →
Website𝕏@RezaRezvaniBlnGitHub

ar-resume

Resumes paused autoresearch experiments by loading branch state and results history.

25.3k|3.6k|Updated Oct 19, 2025
Search Tags:#resume#Optimization#experiment#git branch#autonomous loop#AutoResearch#results history
One-click install
npx skills add https://github.com/alirezarezvani/claude-skills --skill ar-resume
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ar-resume
Source: https://github.com/alirezarezvani/claude-skills/tree/main/engineering/autoresearch-agent/skills/ar-resume
Command: npx skills add https://github.com/alirezarezvani/claude-skills --skill ar-resume

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

When an autonomous research experiment is paused or interrupted by context limits, you lose track of where it left off. This Skill restores the full experiment state so iteration can continue without re-reading everything manually.

Core Features & Use Cases

  • Experiment Discovery: Lists all experiments with status (active, paused, done) derived from results.tsv age when no experiment is specified.
  • Full Context Restoration: Checks out the experiment branch, reads config.cfg, program.md, results.tsv, and recent git history.
  • State Summary & Handoff: Reports experiment counts, best metric, recent patterns, and offers next actions (single run, autonomous loop, or review).
  • Use Case: You paused an API latency optimization experiment yesterday. Run /ar:ar-resume engineering/api-speed to see 23 experiments, the best result of 185ms, and pick up with a new iteration immediately.

Quick Start

Ask the agent to resume the paused autoresearch experiment for your target, for example by running /ar:ar-resume engineering/api-speed.

Frequently Asked Questions about ar-resume

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

FAQPage Schema
How do I resume a paused autoresearch experiment?▼

Run /ar:ar-resume with the experiment path, such as /ar:ar-resume engineering/api-speed. The skill checks out the experiment branch, loads config, strategy, and results history, then summarizes the current state and asks how to continue.

How do I list all autoresearch experiments and their status?▼

Run /ar:ar-resume without arguments. It executes setup_experiment.py --list and shows each experiment as active, paused, or done based on the age of its results.tsv file, letting you pick one to resume.

What state does ar-resume restore when resuming an experiment?▼

It checks out the autoresearch/{domain}/{name} git branch, then reads config.cfg, program.md, the full results.tsv history, and the last 20 git log entries to rebuild complete experiment context.

Can I continue an experiment in autonomous loop mode after resuming?▼

Yes. After the state summary, you can choose to start a loop, which hands off to /ar:loop with the experiment pre-selected, or run a single iteration via /ar:run.

What happens if the experiment branch or results file is missing?▼

The resume flow depends on the autoresearch branch and .autoresearch directory existing. If they are missing, the git checkout or file reads will fail, so the experiment must have been properly initialized beforehand.

Related Skills

Explore More Software Engineering Skills→
Elijah-JElijah-J

ar-start

Initialize autoresearch loops for prepared benchmark projects.

Community
Advanced
ano4lano4l

autoresearch-agent

Automate single-file optimization experiments with git-based keep/discard lifecycle.

Community
Advanced
xlelibertyxleliberty

resume

Restore paused research experiments by checking out git branches and loading historical data.

Community
Intermediate
joshsmithxrmjoshsmithxrm
5

reset

Delete state.json to reset corrupted workflow state for the current branch.

Community
Basic