AutoDream Memory

PassAudited by ClawScan on May 10, 2026.

Overview

The skill’s behavior matches its memory-cleanup purpose, but enabling it lets it read session history and periodically rewrite local memory files.

Install only if you want automatic local memory maintenance. Before enabling the schedule, try --dry-run, review the generated reports and backups, and remember that session history and MEMORY.md may contain sensitive information that could be retained or used in future agent context.

Findings (2)

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

Selected past-session content may be reflected in local memory reports or future memory updates, which can shape later agent behavior.

Why it was flagged

The skill scans local session logs and captures message snippets as signals for memory consolidation. This is central to its stated purpose, but session history can contain private or sensitive context and may influence future memory.

Skill content
sessions_dir = workspace / "sessions" ... for jsonl_file in sessions_dir.glob("*.jsonl") ... "text": text[:200]
Recommendation

Run a dry run first, review generated memory/autodream reports and backups, and avoid keeping secrets in session or memory files.

What this means

If scheduled, the skill can continue reorganizing memory after the initial setup without prompting each time.

Why it was flagged

The helper creates an enabled recurring-run configuration. This persistence is disclosed and aligned with the AutoDream purpose, but users should knowingly enable and manage it.

Skill content
"command": f"python3 skills/autodream/scripts/autodream_cycle.py --workspace .", "interval": f"{interval_hours}h", "enabled": True
Recommendation

Only run the scheduling setup if you want automatic memory cleanup, and remove or disable the OpenClaw cron entry if you no longer want it.