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 agentmods add skills/rexleimo/aios/harness-init-runnernpx skills add rexleimo/aios --skill harness-init-runnergit clone --depth 1 https://github.com/rexleimo/aiosWrote 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/rexleimo/aios/harness-init-runner)<a href="https://agentmods.dev/skills/rexleimo/aios/harness-init-runner"><img src="https://agentmods.dev/badge/skills/rexleimo/aios/harness-init-runner.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00070 | $0.00782 |
| Opus 5 | $0.00035 | $0.00391 |
| Sonnet 5 | $0.00014 | $0.00156 |
| Haiku 4.5 | $0.00007 | $0.00078 |
Grade A, and why
harness-init-runner 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Harness Init Runner (Node.js)
Working directory: any Node.js repo with package.json
Create a portable long-running-agent harness inside the current repo (Node.js), without pulling in the full AIOS workspace layout.
When to use (even if AIOS is installed)
- Use this when you want a repo-local, lightweight
harness/runner that works across Codex / Claude Code / Gemini / opencode. - If you already use AIOS, you may not need this. However it can coexist: this skill creates
harness/+/.harness/and does not depend onscripts/aios.mjs.
What this skill generates (repo root)
harness/(runner code)harness.config.json(provider command templates)/.harness/(runtime artifacts root; gitignored)package.jsonscripts:harness:run,harness:doctor- dependency:
zod
Runtime artifacts are written under ./.harness/runs/* and MUST NOT be committed.
Preconditions
- Run from (or inside) a Node.js repo that has a
package.json. - If
package.jsonis missing, stop and ask the user whether to create a Node project first.
Init Steps (deterministic)
- Locate repo root by searching upward for
package.json. - Copy the bundled templates from
assets/template/into the target repo root:- copy
assets/template/harness/→<repoRoot>/harness/ - For
harness.config.json: if the file already exists, merge (preserve user's existing keys, add only missing ones). If it does not exist, copy the template. Never silently overwrite user edits.
- copy
- Append
/.harness/to<repoRoot>/.gitignore(create file if missing). - Update
<repoRoot>/package.json(additive only):- Add scripts:
harness:run:node harness/run.mjsharness:doctor:node harness/doctor.mjs
- Add dependency
zod(usedependenciesunless the repo clearly wantsdevDependencies).
- Add scripts:
- Run
npm install. - Verify:
npm run harness:doctornpm run harness:run -- --provider codex --task "hello harness"
Safety defaults
What ships with it
14 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.
- assets/template/gitignore-snippet.txt 12 B
- assets/template/harness.config.json 640 B
- assets/template/harness/config.schema.mjs 1.2 KB runs code
- assets/template/harness/doctor.mjs 2.1 KB runs code
- assets/template/harness/lib/checkpoint.mjs 1.4 KB runs code
- assets/template/harness/lib/human-gate.mjs 5.4 KB runs code
- assets/template/harness/lib/io.mjs 302 B runs code
- assets/template/harness/lib/paths.mjs 1.1 KB runs code
- assets/template/harness/providers/claude.mjs 144 B runs code
- assets/template/harness/providers/codex.mjs 142 B runs code
- assets/template/harness/providers/gemini.mjs 144 B runs code
- assets/template/harness/providers/index.mjs 425 B runs code
- assets/template/harness/providers/opencode.mjs 148 B runs code
- assets/template/harness/run.mjs 6.8 KB runs code
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.
- 4d ago First seen · 68 lines · 70 tokens per session scan A d0face01ef96
harness-init-runner is a skill published in the GitHub repository rexleimo/aios (52 stars, last pushed today), licensed MIT. It adds 70 tokens to every session and 782 once invoked, about $0.0003 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.
Other skills, from other repositories
code-review
对代码变更进行全面的代码审查，检查质量、安全性和最佳实践.
hello-world
一个简单的演示 Skill，验证 Agent Skill 系统的基本功能.
project-report
生成项目报告：分析目录结构、统计代码量、识别关键文件.
ralph
Interactive task planning and autonomous execution using @open-harness/ralph. Interviews user, creates PRD, generates task list with Claude Code's built-in task tools, then executes tasks autonomously with full Claude Code capabilities. USE WHEN user says 'ralph', 'run ralph', 'autonomous task', 'execute plan'…
arxiv
Standardized arXiv skill for MediaUse. Includes Windows install, key onboarding, strict context/auth flow, full arXiv dynamic command map, and safety/rate controls.
douyinv2
Standardized douyinv2 skill for MediaUse. Includes Windows install, key onboarding, strict context/auth flow, full douyinv2 dynamic command map, and safety/rate controls.