skills.rest
googlarz
googlarzCOMMUNITY
Explore all skills by this author →
GitHub

terminal-bench-loop

Coordinate bounded Terminal-Bench smoke loops with diagnostics and board-approved fixes.

3|Updated Apr 11, 2026
Search Tags:#workflow#worktree#diagnosis#loop#terminal-bench#paperclip
One-click install
npx skills add https://github.com/googlarz/stapler --skill terminal-bench-loop
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: terminal-bench-loop
Source: https://github.com/googlarz/stapler/tree/main/skills/terminal-bench-loop
Command: npx skills add https://github.com/googlarz/stapler --skill terminal-bench-loop

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinates a bounded, human-in-the-loop process to drive Terminal-Bench problems through Paperclip until a smoke passes or blockers are named. It captures artifacts, diagnoses the exact stop point with /diagnose-why-work-stopped, and routes a board-confirmation for any product fix, then reruns against the same worktree.

Core Features & Use Cases

  • Bounded iteration governance: enforces a maximum number of iterations and per-iteration time limits.
  • Artifact capture & diagnosis: records run IDs, manifests, results, and the stop reason, and applies /diagnose-why-work-stopped within the loop subtree.
  • Isolated worktree discipline: runs against a single Paperclip App worktree with workspace inheritance across iterations.
  • Decision routing: gates product fixes behind board confirmation and manages blockers with explicit ownership.

Quick Start

Define the loop inputs (source issue, task name, budget, and workspace), then start the first bounded smoke against the isolated Paperclip App worktree.

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 repetitive Terminal-Bench smoke tests with a bounded iteration limit?▼

Automated Terminal-Bench smoke tests can be looped with explicit diagnostic capture by defining source issues, task names, budgets, and workspace pointers to enforce maximum iteration limits and per-iteration time constraints.

Why does my Terminal-Bench workflow stop running and how do I diagnose the exact stop point?▼

Diagnosing Terminal-Bench work stoppage is handled by applying /diagnose-why-work-stopped within the loop subtree, capturing run IDs, manifests, results, and the exact stop reason to trace blockers.

Can I gate product fixes behind board confirmation during Terminal-Bench diagnosis loops?▼

Product fixes in Terminal-Bench loops are gated behind board confirmation, routing decisions for approval and managing explicit blockers with ownership before rerunning against the same worktree.

What inputs do I need to set up a bounded Terminal-Bench looping process?▼

Setting up bounded Terminal-Bench looping requires defining source issues, task names, budget constraints, worktree pointers, artifact capture settings, and a diagnosis pattern to ensure traceable blockers and approval workflows.

Related Skills

Explore More Software Engineering Skills→
bychrisrbychrisr

terminal-bench-loop

Orchestrate bounded human-in-the-loop improvement cycles for Terminal-Bench problems.

Community
Advanced
agentic-inagentic-in
577

Codex

Delegate bounded coding tasks to Codex for isolated worktree patches.

Official
Intermediate
dotlab-hqdotlab-hq
1

playwright

Automate browser navigation, form filling, and data capture from the terminal.

Official
Intermediate
o9nno9nn

terminal-bench-loop

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

Official
Advanced