Restart Task Recovery
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This skill is transparent and purpose-aligned for restart recovery, but it stores session context and can send checkpoint-derived resume messages to other agents.
This appears safe for its stated purpose, but before installing or using it, make sure you are comfortable with it saving concise task checkpoints and sending resume messages to other agents after a restart. Review generated recovery JSON/checkpoints before execution, especially for tasks that might write data, publish content, deploy changes, or repeat external actions.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
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.
Other agents may be prompted to continue prior tasks after a restart.
The skill directs the agent to use session-control tooling to resume other active sessions. This is central to restart recovery and includes limits such as no tight loops, but users should recognize it can cause other agents to continue work.
For each interrupted/idle target session, send resume message via `sessions_send`
Review the generated recovery actions before sending them, especially if the interrupted work involved external writes, publishing, deployments, or account changes.
Task goals, recent progress, blockers, and resume messages may remain stored locally after recovery.
The skill persists recent session context and resume instructions so they can be reused after restart. This is purpose-aligned, but checkpoint files may contain sensitive task details or stale instructions.
Write a compact checkpoint file at: `memory/restart-checkpoints/<YYYY-MM-DD>/<HHmmss>.md`
Keep checkpoints concise, avoid including secrets, review checkpoint contents before reuse, and delete old checkpoints when no longer needed.
If a checkpoint is inaccurate or overly broad, the wrong session could receive an incorrect resume instruction.
The helper converts checkpoint entries into session-send actions. This is expected for the skill, but it means checkpoint content directly shapes messages sent to other agents.
"actions": [{"tool": "sessions_send", "params": {"sessionKey": i["sessionKey"], "message": i["message"]}} for i in items]Compare the generated action list against the active session list and confirm session keys/messages before execution.