terminal-bench-loop

Run Terminal-Bench tasks in bounded loops with diagnostic stop analysis.

Updated May 4, 2026
One-click install
npx skills add https://github.com/o9nn/entelechorg --skill terminal-bench-loop-o9nn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: terminal-bench-loop
Source: https://github.com/o9nn/entelechorg/tree/main/skills/terminal-bench-loop
Command: npx skills add https://github.com/o9nn/entelechorg --skill terminal-bench-loop-o9nn

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill streamlines the process of executing and managing Terminal-Bench problem loops, ensuring bounded, diagnostic, and human-in-the-loop control over iterative testing workflows.

Core Features & Use Cases

  • Automated Loop Execution: Runs Terminal-Bench problems repeatedly with bounded iterations, verifying smoke pass/fail status.
  • Diagnostic Analysis: Applies /diagnose-why-work-stopped to identify stop reasons at each iteration.
  • Controlled Fix Proposals: Coordinates product fix proposals and requests board confirmation before implementation, supporting safe iterative development.

Quick Start

Run Terminal-Bench problems in a bounded loop with diagnostic support to verify stability and identify issues efficiently.

Frequently Asked Questions about terminal-bench-loop

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

FAQPage Schema
How do I automate Terminal-Bench testing with bounded loop iterations?

Automate Terminal-Bench testing by running problems repeatedly with bounded iterations, verifying smoke pass/fail status at each cycle to ensure reliable problem-solving workflows.

How does diagnostic analysis work during automated testing loops?

Diagnostic analysis applies automated stop-reason identification at each loop iteration, pinpointing why work stopped so you can request board confirmation before implementing controlled fixes.

Can I require human approval before applying fixes in an automated testing workflow?

Yes, the workflow coordinates product fix proposals and requests board confirmation before implementation, ensuring human-in-the-loop control over safe iterative development.

What is the best way to identify why a Terminal-Bench task stopped executing?

Apply diagnostic stop-reason analysis to identify why work stopped at each iteration, verifying stability and surfacing issues efficiently within a bounded testing loop.

Do I need any external dependencies to run bounded diagnostic testing loops?

No external dependencies are required; the workflow uses internal scripts and references to coordinate bounded iterations, automated diagnosis, and human-in-the-loop approval processes.

When should I use a bounded testing loop instead of unbounded test execution?

Use bounded loops when you need controlled, iterative testing with diagnostic analysis and safe approval processes, preventing unbounded execution from running indefinitely without stability verification.