Heartbeat Checklist

WarnAudited by ClawScan on May 10, 2026.

Overview

This skill asks the agent to periodically change your computer and long-term memory, including closing apps, cleaning tabs, trashing screenshots, restarting services, and editing memory files without clear confirmation limits.

Review carefully before installing. Use it only if you want an agent to perform recurring maintenance, and require confirmations for all system changes, file moves, app/tab closures, and memory edits.

Findings (3)

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.

What this means

The agent could disrupt work, lose browser state, move files, or change background services during routine checks.

Why it was flagged

The skill instructs the agent to run local system commands and mutate the user's environment, including service/timer changes, app/tab closure, and moving files to trash, without a clear approval requirement.

Skill content
`hydra status` - restart if inactive ... `systemctl --user is-active resource-monitor.timer` - if dead, reenable ... Close unused apps ... Clean stale browser tabs ... Move old screenshots to trash
Recommendation

Require explicit user confirmation before any restart, reenable, app/tab closure, or file move, and define exactly which files, apps, and services are in scope.

What this means

The agent may repeatedly perform maintenance actions outside a specific user request.

Why it was flagged

The skill directs recurring autonomous behavior, but the artifact does not define a controlled schedule, opt-in mechanism, stop condition, or audit trail for repeated self-healing actions.

Skill content
Run these checks periodically (2-4x daily).
Recommendation

Make the schedule opt-in, visible, and easy to disable; log each run and ask before making changes.

What this means

Sensitive or incorrect session details could be stored and reused later, while existing memory could be changed or removed without review.

Why it was flagged

The skill writes summaries and edits persistent memory files, but does not bound what content may be saved, how sensitive data is handled, or when the user approves memory changes.

Skill content
If > 60%: Save summary to `memory/YYYY-MM-DD.md` ... Review recent daily notes ... Update MEMORY.md with distilled learnings ... Remove outdated info
Recommendation

Ask before writing or deleting memory, show the proposed content, exclude secrets/private data, and keep reversible history for memory edits.