← Back to Skills Marketplace
50
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install workload-scheduler
Description
Selects appropriate workload templates from the pool based on available system capacity
README (SKILL.md)
负载调度器
概述
接收容量监控器的空闲容量报告，从负载模板池中选择适合当前容量的任务进行调度。优先选择优先级高且资源需求匹配的模板，最大化资源利用率。
使用场景
- 资源利用最大化：在空闲容量充足时运行辅助任务
- 智能排程：基于可用资源自动选择合适的工作负载
- 负载管理：确保不超过系统容量上限
调度规则
- 读取所有负载模板及其资源需求。
- 过滤出资源需求不超过空闲容量的模板。
- 按优先级和资源匹配度排序选择。
- 生成调度计划。
使用方法
- 接收容量监控报告获取空闲容量。
- 从模板目录读取负载模板配置。
- 过滤和排序可选负载模板。
- 输出调度计划 JSON。
输出格式
{
"schedule_timestamp": "2026-06-01T12:00:05",
"capacity": {"cpu_percent_idle": 75.0, "memory_mb_available": 4096},
"scheduled_tasks": ["log-rotate", "analytics-aggregation"],
"status": "completed"
}
实现代码
见 scripts/schedule.py。
Usage Guidance
Install only if you want agent help with ClawHub/Convex development and maintenance workflows. Before using the moderation, proof publishing, migration, or autoreview helper paths, review the exact command because they can modify services, publish PR artifacts, run local project code, or send diffs to configured reviewer CLIs; use the autoreview --no-yolo option if you do not want the nested review to bypass sandbox approvals.
Capability Assessment
Purpose & Capability
The skills cover Convex setup, auth, migrations, performance review, ClawHub PR maintenance, UI proof, moderation, and autoreview; higher-impact actions such as moderation changes, migrations, proof publishing, and code review commands are aligned with those stated purposes.
Instruction Scope
Runtime instructions are mostly scoped and user-directed, including explicit target and confirmation requirements for moderation writes; the autoreview helper discloses that it defaults to a full-access nested Codex review and provides a --no-yolo opt-out.
Install Mechanism
The inspected skill files and agent YAML do not show hidden install hooks, obfuscated setup, or automatic persistence; package-level scripts seen in the workspace are ordinary development/build/test commands.
Credentials
Use of gh, git, bun/npm/npx, Convex, local dev servers, and external reviewer CLIs is proportionate for the documented maintainer, proof, setup, and review workflows, but users should expect local code execution and external service access when invoking those paths.
Persistence & Privilege
No hidden background persistence was found; long-running Convex/dev proof processes are disclosed as workflow steps, and elevated review privileges are disclosed rather than concealed.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install workload-scheduler - After installation, invoke the skill by name or use
/workload-scheduler - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of workload-scheduler.
- Selects workload templates based on available system capacity.
- Prioritizes tasks by resource requirements and priority to maximize utilization.
- Automatically creates a schedule plan in JSON output.
- Designed for scenarios needing resource optimization and intelligent workload management.
Metadata
Frequently Asked Questions
What is Workload Scheduler?
Selects appropriate workload templates from the pool based on available system capacity. It is an AI Agent Skill for Claude Code / OpenClaw, with 50 downloads so far.
How do I install Workload Scheduler?
Run "/install workload-scheduler" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Workload Scheduler free?
Yes, Workload Scheduler is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Workload Scheduler support?
Workload Scheduler is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Workload Scheduler?
It is built and maintained by InkLin (@ink5725); the current version is v1.0.0.
More Skills