meta-harness-terminal-bench-2

meta-harness-terminal-bench-2 is a skill for Claude Code from stanford-iris-lab/meta-harness. It costs 23 tokens per session (1,662 once invoked), scanned A, original, MIT.

A skill for improving an agent setup through one iteration of AgentHarness evolution for Terminal-Bench 2, a benchmark of command-line software tasks. It analyzes results and failed attempts, then produces a new general-purpose agent variant.

In plain words
What is it for?
Use it to review benchmark trajectories, identify broadly useful improvements, and implement one updated agent scaffold per iteration.
Why use it?
It provides a controlled way to improve an agent based on observed failures without adding hints for individual benchmark tasks. Each iteration must result in a new variant.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions subagents.

Good fit Use it to review benchmark trajectories, identify broadly useful improvements, and implement one updated agent scaffold per iteration.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2
About the project

Meta-Harness is a framework that automatically searches for task-specific harnesses, the surrounding code that controls what a fixed model stores, retrieves, and displays during work. Researchers and developers use it to optimize agent workflows, with reference experiments for text classification and terminal tasks; catalogue skills are examples of applying this approach.

stanford-iris-lab/meta-harness · 1,549 stars · on GitHub · yoonholee.com

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add stanford-iris-lab/meta-harness --skill meta-harness-terminal-bench-2
Clone the repo
git clone --depth 1 https://github.com/stanford-iris-lab/meta-harness

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for meta-harness-terminal-bench-2

README.md
[![agentmods](https://agentmods.dev/badge/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2/github.svg)](https://agentmods.dev/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2)
Your own site
<a href="https://agentmods.dev/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2"><img src="https://agentmods.dev/badge/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for meta-harness-terminal-bench-2

Your own site · 80×15
<a href="https://agentmods.dev/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2"><img src="https://agentmods.dev/badge/skills/stanford-iris-lab/meta-harness/meta-harness-terminal-bench-2.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 23 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,662 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00023 $0.01662
Opus 5 $0.00012 $0.00831
Sonnet 5 $0.00005 $0.00332
Haiku 4.5 $0.00002 $0.00166

Measured 12d ago against content hash 37d148f046ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

meta-harness-terminal-bench-2 scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

reference_examples/terminal_bench_2/.claude/skills/meta-harness-terminal-bench-2/SKILL.md · 130 lines

How it starts

The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Meta-Harness (Terminal-Bench 2)

Run ONE iteration of agent scaffold evolution.

You do NOT run benchmarks. You analyze results + failed trajectories, propose agent variants, and implement them. The outer loop (meta_harness.py) handles benchmarking.

CRITICAL CONSTRAINTS

  • You MUST produce 1 new agent variant every iteration
  • Do NOT write "the frontier is optimal" or "stop iterating", or abort early.

Anti-overfitting rules

  • No task-specific hints. Do not hardcode knowledge about specific tasks. Agents must be general-purpose.
  • Never mention task names in agent code, prompts, or comments. No references like "if task contains 'async'" or "for polyglot tasks." If your improvement only helps one task, it's too specific.
  • General guidance is OK. Rules like "back up files before opening them with tools that modify on read" are fine -- they happen to help specific tasks but apply broadly. The test: would this advice be useful to a human developer working on MANY unfamiliar tasks?
  • If in doubt, make it more general. "Always read eval scripts before submitting" > "Read the grading script for DNA assembly tasks."

CONTEXT

You are evolving the AgentHarness agent scaffold for Terminal-Bench 2. It is located in agents/baseline_kira.py.

The search space is arbitrary Python code. You can override any method, call any library, make raw API calls, add new tools, change how the LLM is called, rewrite command execution, intercept and transform observations -- anything that's expressible in Python is fair game. The only constraint is that the agent must subclass harbor.agents.terminus_2.terminus_2.Terminus2 in the same way as baseline_kira.py does (for compatibility with the eval harness).

Model: Claude Opus 4.6. Evaluation uses the full official 89-task TB2 dataset. meta_harness.py chooses the trial count; the default path uses 2 trials per task.

Key files to read:

  • agents/baseline_kira.py - the full baseline implementation. Read to understand overridable methods.

Read the full file on GitHub · 130 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 130 lines · 23 tokens per session scan A 37d148f046ac

Subscribe to this mod's changes

meta-harness-terminal-bench-2 is a skill published in the GitHub repository stanford-iris-lab/meta-harness (1,549 stars, last pushed today), licensed MIT. It adds 23 tokens to every session and 1,662 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

building

Implementation skill for writing production code with TDD. Covers the RED-GREEN-REFACTOR cycle, false-RED detection, vertical slicing, scope escalation, test process discipline, and code generation patterns. Loaded by component-builder and bug-investigator.

romiluz13/cc10x · 53 tokens

diff-driven-docs

Use when a BUILD phase completes, a commit is staged, or a PR is about to be created, and the diff has not yet been reflected in documentation. Also use when the user says "update docs", "sync docs", "document this", or asks whether documentation is up to date.

romiluz13/cc10x · 64 tokens

exploration

Two-mode exploration skill: (1) design dialogue — turn rough ideas into validated designs through collaborative interview before planning; (2) spike — throwaway code answering ONE design question, deleted or absorbed, never shipped. Router invokes mode via dispatch context.

romiluz13/cc10x · 56 tokens

memory-and-handoff

Two-mode skill: (1) session memory — load/persist durable workflow state under .cc10x/ (activeContext, patterns, progress) so context survives compaction; (2) handoff package — portable, secrets-redacted export for a coworker, different tool, or fresh non-cc10x session.

romiluz13/cc10x · 70 tokens

plan-review-gate

Use after saving a non-trivial plan or decision RFC when a fail-closed feasibility, completeness, and alignment review must block execution.

romiluz13/cc10x · 32 tokens

architecture

Greenfield architecture design: map functionality flows, draw components, design APIs, classify dependencies, plan observability. For multi-component, API, schema, auth, or integration-heavy work. For retrofitting existing code, use codebase-hygiene instead.

romiluz13/cc10x · 56 tokens