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/painhardcore/pstack/arenanpx skills add painhardcore/pstack --skill arenagit clone --depth 1 https://github.com/painhardcore/pstackWhat 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 | $0.00058 | $0.01041 |
| Opus 5 | $0.00029 | $0.00521 |
| Sonnet 5 | $0.00012 | $0.00208 |
| Haiku 4.5 | $0.00006 | $0.00104 |
Grade A, and why
arena 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 yesterday.
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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Arena
Fan out N parallel attempts at the same task. Read every candidate end to end. Pick the strongest as the base. Graft the best ideas from the others into it. Verify the synthesized result.
Start
Open a todolist with one entry per phase before launching anything. The arena runs autonomously and the list keeps phases from silently disappearing.
- Frame
- Fan out
- Cross-judge
- Pick
- Graft
- Verify
Phase A: Frame
The N candidates will receive the same prompt, so the prompt is the contract. Get it right before spawning anything.
- State the artifact each candidate is producing.
- Derive the rubric. State what success looks like for this task, then turn it into 3-6 concrete gradeable criteria. Concrete:
Adds a --dry-run flag that skips writes. Vague:code is correct. The rubric is the picker's tool in Phase D; candidates only see the task. - Pick the runners. Use the parent model by default. When the host supports model selection, prefer different available model families for judgment-heavy work. Reusing the parent model is fine when generation breadth matters more than model diversity.
- Assign output paths. Each candidate writes to its own location (a git worktree where possible, otherwise
/tmp/arena-<slug>/candidate-<n>/). N candidates writing to the same path is shared mutable state and fails the the separate-before-serializing-shared-state principle skill test.
Phase B: Fan out
Use the host's native subagent tool to start all N candidates concurrently. Give each candidate the task, the shared grounding path, its own output path, and instructions to produce both the artifact and a short rationale. If subagents are unavailable or forbidden, produce the candidates sequentially in the same isolated locations before judging them.
The rationale is mandatory. Without it, the parent cannot tell whether a candidate's structure is principled or accidental, which makes Phase E grafting unreliable. Each rationale names the alternatives the candidate considered and what it rejected.
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.
- yesterday First seen · 71 lines · 58 tokens per session scan A 307cc4950210
arena is a skill published in the GitHub repository painhardcore/pstack (1 stars, last pushed 5d ago), licensed MIT. It adds 58 tokens to every session and 1,041 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 2026-08-31.
Other skills, from other repositories
skill-authoring
Author SKILL.md skills: frontmatter, validator limits, structure.
iflytek-hyper-tts
Use when user asks to synthesize speech, convert text to audio, or read text aloud. 讯飞超拟人语音合成 - 支持文本转语音、语音合成（发音人/语速/语调/音量/输出格式）。大模型语音合成技能。语音合成, 文字转语音, 超拟人, TTS.
iflytek-ocr-invoice
Use when user asks to recognize invoices, extract receipt data, or OCR bills and tickets. Recognize and extract structured data from invoices, receipts, and bills using iFlytek OCR API (科大讯飞票据识别). Supports VAT invoices, taxi receipts, train tickets, toll invoices, medical bills, bank receipts, and more.
iflytek-pdf-image-ocr
AI-powered OCR service for images and PDF documents using iFlytek's advanced recognition APIs.
iflytek-voiceclone-tts
Use when user asks to clone a voice, train a custom voice model, or synthesize speech with a cloned voice. iFlytek Voice Clone tts(声音复刻) — train a custom voice model from audio samples and synthesize speech with the cloned voice. Supports the full workflow: get training text → create task → upload audio → submit…
animated-sketch-diagram
生成"黑墨手绘涂鸦"风格的动画架构图/流程图：米色纸面、针管笔墨线、极淡水洗色块、简笔涂鸦图标、序号章、连线上的流动圆点动画、图标微动效。产出单文件自包含动画 HTML（SVG+CSS），可一键导出无缝循环 GIF。当用户想画架构图、流程图、信息图、技术示意图、对比图、pipeline/workflow 可视化，或提到"手绘风""涂鸦风""动图""animated diagram""GIF 架构图"时使用；即使用户没明说要动画，做技术概念科普配图时也应优先考虑本 skill。.