LongHorizon-Harness is a computer-use harness that lets AI agents continue work across desktop applications and the command line for extended periods by planning, acting, verifying, checkpointing, and recovering. It is for users who need Claude Code, Codex, OpenCode, or DeepSeek Harness to make reliable progress on complex long-running workflows without training a new model. The catalogue entries provide skills for operating this execution loop.
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.
npx skills add AMAP-ML/LongHorizon-Harness --skill setup-osworldgit clone --depth 1 https://github.com/AMAP-ML/LongHorizon-HarnessWrote 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.
[](https://agentmods.dev/skills/amap-ml/longhorizon-harness/setup-osworld)<a href="https://agentmods.dev/skills/amap-ml/longhorizon-harness/setup-osworld"><img src="https://agentmods.dev/badge/skills/amap-ml/longhorizon-harness/setup-osworld/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.
<a href="https://agentmods.dev/skills/amap-ml/longhorizon-harness/setup-osworld"><img src="https://agentmods.dev/badge/skills/amap-ml/longhorizon-harness/setup-osworld.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Privilege Escalation · line 80 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00101 | $0.01102 |
| Opus 5 | $0.00051 | $0.00551 |
| Sonnet 5 | $0.00020 | $0.00220 |
| Haiku 4.5 | $0.00010 | $0.00110 |
Grade B, and why
setup-osworld scanned grade B with 1 finding 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 11d 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.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
- secret exports in a chmod 600 local file unless the user explicitly asks How it starts
The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Setup OSWorld
Use this skill to make a cloned OSWorld-V2 checkout runnable, not merely to list environment variables. Provision every selected component that can be configured safely from the current machine, and ask the user whenever an action requires credentials, paid resources, DNS changes, SSH keys, secrets, or a destructive/cloud operation.
Supported Scope
Supported provider setup:
aws: fully provision required AWS network resources with AWS CLI, then export the values OSWorld needs.docker: install/verify Docker and KVM where possible, then verify the Docker provider can run.
Unsupported provider setup for now:
vmware,virtualbox,azure,gcp,aliyun, andvolcengine.
For unsupported providers, load references/unsupported-providers.md, state
that this skill does not automate them yet, and ask whether the user wants AWS
or Docker instead.
Optional service setup:
- OSWorld-web mocked website server, using
Task-Web/OSWorld-web. - GitLab server, using
Task-Web/gitlab. - Gated task class download from Hugging Face.
- Task-scoped proxy setup following upstream OSWorld section 2.3, plus optional host-side proxy or HF mirror setup.
- CUA-Harness hybrid experiment overlay when this checkout includes
experiments/osworld_v2_hybridand the siblingcua-harness/source tree.
Start With Intake
Ask before doing infrastructure work unless the user has already supplied the answers.
Ask for:
- provider:
awsordocker - whether to use existing AWS resources or create new ones
- AWS region, VPC/subnet preference, and whether AWS charges are acceptable
- whether to set up OSWorld-web, and on existing server or new AWS EC2 host
- website domain mode: existing wildcard domain,
nip.io, or user-provided host suffix - whether to set up GitLab, and whether it must be a separate server/domain
- whether to download gated task classes, latest or a benchmark release
- whether task-scoped proxy, host-side proxy, or HF mirror is needed
- whether to configure the CUA-Harness hybrid experiment overlay
- whether to run real smoke tests that create/start resources
What ships with it
9 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 11d ago First seen · 104 lines · 101 tokens per session scan B 97bc6a39ed81
setup-osworld is a skill published in the GitHub repository AMAP-ML/LongHorizon-Harness (1,481 stars, last pushed 21d ago), licensed MIT. It adds 101 tokens to every session and 1,102 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
open-code-review
Performs AI-powered code review on Git changes using the ocr CLI from alibaba/open-code-review. Use when the user asks to review code, review a pull request, review staged/unstaged changes, review a commit, or compare branches for code quality issues. Produces line-level review comments and can automatically apply…
harness-creator
Build, audit, and improve harnesses that make AI coding agents reliable: AGENTS.md/CLAUDE.md instruction files, feature/state tracking, verification gates, scope boundaries, session handoff, memory persistence, context budgets, tool-permission safety, and multi-agent coordination. Use this whenever a coding agent is…
remote-claude-code
Run Claude Code on a remote host over SSH — a persistent expect-driven login session, headless claude -p with the stdin fix, the interactive TUI inside a remote tmux driven by send-keys/capture-pane (one keystroke at a time, capture-verified; relayed user messages go through verbatim), and multi-turn continuity via…
agent-optimization
Improve an Agent State through versioned scores and score-linked Traces from a frozen Benchmark.
web-design
Penguin visual language for generated web pages and app UIs — GitHub-style simplicity with a single blue accent, light and pure-black dark themes, design tokens, component and chat-interface recipes, plus an opt-in warm paper editorial theme.
agent-evaluation
Run one specified Test Agent on one specified Benchmark Case exactly once, privately score that execution, and return one protocol result.