# What Makes a Good Terminal-Agent Benchmark Task: A Guideline for Adversarial, Difficult, and Legible Evaluation Design

arXiv: https://arxiv.org/abs/2604.28093 · HTML: https://mlsanity.com/paper/2604.28093/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Agent skill learning and evolution, one of 16 directions in LLM agents and tool use.
- Builds on 7 of its 13 references that are in the map.
- First posted on arXiv 30 Apr 2026.

## Placed in

- LLM agents and tool use / [Agent skill learning and evolution](https://mlsanity.com/agents/agent-skill-learning-and-evolution/)

## Builds on (most cited first, 7 of 7)

- [Concrete Problems in AI Safety](https://mlsanity.com/paper/1606.06565/) (2016)
- [Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces](https://mlsanity.com/paper/2601.11868/) (2026)
- [Do the Rewards Justify the Means? Measuring Trade-Offs Between Rewards and Ethical Behavior in the MACHIAVELLI Benchmark](https://mlsanity.com/paper/2304.03279/) (2023)
- [Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models](https://mlsanity.com/paper/2406.10162/) (2024)
- [Natural Emergent Misalignment from Reward Hacking in Production RL](https://mlsanity.com/paper/2511.18397/) (2025)
- [Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem](https://mlsanity.com/paper/2512.24873/) (2025)
- [Terminal Wrench: A Dataset of 331 Reward-Hackable Environments and 3,632 Exploit Trajectories](https://mlsanity.com/paper/2604.17596/) (2026)

## Built on by (most cited first, 3 of 3)

- [Model or Harness? An Interaction-Centric Taxonomy for Localizing Agent Failures](https://mlsanity.com/paper/2607.28802/) (2026)
- [CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks](https://mlsanity.com/paper/2608.06352/) (2026)
- [Environment Evolution for Terminal Agents](https://mlsanity.com/paper/2609.04128/) (2026)

## Abstract (arXiv)

Terminal-agent benchmarks have become a primary signal for measuring the coding and system-administration capabilities of large language models. As the market for evaluation environments grows, so does the pressure to ship tasks quickly, often without thorough adversarial review of the verification logic. This paper is a guideline for writing good benchmark tasks, drawn from over a year of contributing to and reviewing tasks for Terminal Bench. Most people write benchmark tasks the way they write prompts. They shouldn't. A prompt is designed to help the agent succeed; a benchmark is designed to find out if it can. We argue that good tasks are adversarial, difficult, and legible, and that a large class of common failure modes -- AI-generated instructions, over-prescriptive specifications, clerical difficulty, oracle solutions that assume hidden knowledge, tests that validate the wrong things, and reward-hackable environments -- are predictable consequences of treating task authoring as prompt authoring. We catalog these failure modes, argue that real difficulty is conceptual rather than environmental, and discuss recent empirical evidence that over 15% of tasks in popular terminal-agent benchmarks are reward-hackable. We hope this serves as a useful reference for benchmark maintainers, task contributors, and researchers using benchmark scores as evidence.