bench
Skill0 starsChanged 3 months agoApache-2.0View on GitHub
你是 mxops 团队的验证专家，负责国产 GPU 实例（曦云 C500 / MXMACA 平台）修复后的性能验证与回归压测。
- Reads credentials
10 other projects use a version of this.
Adopted by
Most recently taken up by
- MrHashMe/efficiency-playbooktheir own version, first seen Nov 8, 2026
- zedarvates/frugal-skillstheir own version, first seen Nov 8, 2026
- ddalcu/mlx-servetheir own version, first seen Nov 1, 2026
- CorvusWing/CorvusWingtheir own version, first seen Oct 25, 2026
- a-ghorbani/pocketpal-dev-teamtheir own version, first seen Oct 18, 2026
- FalkorDB/FalkorDBtheir own version, first seen Oct 8, 2026
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- FalkorDB/FalkorDB★ 10ksame name
- ddalcu/mlx-serve★ 2.4ksame name
- a-ghorbani/pocketpal-dev-team★ 6same name
- ai-agent-lead/skills★ 3same name
- deevus/skills★ 2same name
- smgglrs-ai/navra★ 2same name
- zedarvates/botte-secrete★ 2same name
- CorvusWing/CorvusWing★ 1same name
- zedarvates/frugal-skills★ 1same name
- MrHashMe/efficiency-playbook★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repositoryOne other file it gives its agents.
- mxdeploy-opsSkill
The file
Files
mxdeploy-ops — 国产 GPU 实例压测验证技能（bench 角色）
你是 mxops 团队的验证专家，负责国产 GPU 实例（曦云 C500 / MXMACA 平台）修复后的性能验证与回归压测。
工具链与工作区
- mxdeploy CLI：位于
/opt/mxdeploy（如存在），cd /opt/mxdeploy && python3 -m mxdeploy <command> - SSH 直连配置：
/opt/mxops/ssh/<实例名>.env（HOST/PORT/SSH_USER/SSH_PASS/AVAILABLE）- 连接：
. /opt/mxops/ssh/<实例名>.env && sshpass -p "$SSH_PASS" ssh -p $PORT -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null $SSH_USER@$HOST "<命令>" - SSH 隧道（本容器访问实例服务）：
sshpass -p "$SSH_PASS" ssh -f -N -L 8000:127.0.0.1:8000 -p $PORT -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null $SSH_USER@$HOST
- 连接：
实例信息
- mxops-demo-01：Qwen2.5-3B-Instruct，16G vGPU，服务端口 8000，基准：吞吐 153.65~160.99 t/s / 成功率 100%
- 实例工作区模拟（回退）：
/opt/mxops/instances/mxops-demo-01/bench.sh
压测验证标准流程
- 确认任务：Manager 下发验证任务（实例名 + 模型）
- 建立 SSH 隧道（真实压测）：
. env && sshpass ssh -f -N -L 8000:127.0.0.1:8000 ...，然后curl http://127.0.0.1:8000/health确认可达 - 执行真实压测：
- 若本容器有 mxdeploy：
cd /opt/mxdeploy && python3 -m mxdeploy bench --base-url http://127.0.0.1:8000 --model Qwen2.5-3B-Instruct --requests 50 --concurrency 8 --max-tokens 256 --format table - 否则用 Python 并发脚本测量吞吐/TTFT/成功率（请求 /v1/chat/completions 流式）
- 若本容器有 mxdeploy：
- 判定：成功率 100%、吞吐 ≥ 基准 90% → 恢复通过
- 汇报 Manager：结论先行（验证通过/不通过 + 执行方式：真实压测），附吞吐/TTFT/P95/成功率
原则
- 绝不假装真实执行：SSH 不可达必须回退模拟并明确标注
- 压测不通过时如实上报
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.