长任务监控方案。实现 Worker-Monitor 架构，Monitor 通过 hook-logger 日志监控 Worker 状态，每轮 10 分钟通过 Announce 汇报。采用主会话轮询机制（因子代理 sessions_send 限制）。推荐 OpenClaw 2.21+。触发词：长任务、监控任务、任务监控
PassAudited by ClawScan on May 10, 2026.
Overview
The skill is purpose-aligned for monitoring long-running OpenClaw tasks, but it does so by running background sessions, reading local hook logs, and relying on an external logger plugin.
Before installing, make sure you trust the hook-logger plugin, understand that worker/monitor sessions can continue running in the background, and are comfortable with the monitor reading local OpenClaw logs. Manually verify cleanup if a task finishes, especially because the skill documents a current cleanup bug.
Findings (5)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The skill can manage child agent sessions for you, including stopping them during cleanup.
The skill needs authority to create, message, and kill OpenClaw sessions, which is expected for a long-task monitor but is still delegated control over agent sessions.
"permissions": ["sessions-spawn", "sessions-send", "sessions-kill"]
Use it only for tasks where you are comfortable letting the agent create and manage worker/monitor sessions, and keep track of generated session keys.
The logger plugin will affect what monitoring data is available and may record local agent activity.
The skill depends on an external hook-logger plugin that is not included in the reviewed files and is not pinned to a specific reviewed version.
openclaw plugins install @scotthuang/hook-logger
Install hook-logger only from a trusted source and review or pin the plugin version if you use this in sensitive environments.
Installing and using the skill allows local helper commands to run as part of task management.
The helper script executes local commands. In the visible code, this is used for node task-manager calls and OpenClaw session cleanup with argument arrays, which fits the stated purpose.
execFile(cmd, args, { encoding: 'utf-8', ...options },Review commands before running them, especially cleanup commands, and avoid using the skill for untrusted task descriptions.
Monitor reports may include or be influenced by local agent activity logs.
The monitor reads local hook-logger logs and uses them to decide worker status; those logs may contain sensitive task activity or misleading entries if logs are polluted.
读取 hook-logger 日志 文件: ~/.openclaw/workspace/logs/hook-logger/YYYY-MM-DD.log
Use this only when you are comfortable with the monitor reading hook-logger logs, and avoid running it on tasks that may place secrets in logs.
Background worker or monitor sessions may keep running until the task is completed or manually cleaned up.
The skill intentionally creates long-running worker/monitor agents and can continue monitoring automatically for several rounds before asking the user.
创建 Worker 执行长任务 ... 轮次 1-5 | 自动继续下一轮监控
Watch active sessions, use the documented complete/cleanup flow, and manually kill leftover sessions if the documented cleanup bug occurs.