Better Memory

Security checks across static analysis, malware telemetry, and agentic risk

Overview

Better Memory appears purpose-aligned for local workspace memory management, but it creates persistent memory and maintenance artifacts that users should review before enabling automation.

Install this only in the workspace where you want persistent memory. After running smart_install, review AGENTS.md, MEMORY.md, .openclaw-memory-os, optional HEARTBEAT.md, migration/cleanup plans, and any cron template before enabling scheduled maintenance.

Static analysis

No static analysis findings were reported for this release.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

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.

What this means

Incorrect, sensitive, or maliciously phrased memory entries could persist and influence later agent behavior.

Why it was flagged

The skill stores user-provided memory and rolls it into persistent files that may be reused as future agent context.

Skill content
L1 (native): `memory/YYYY-MM-DD.md` ... L2 (sidecar): `.openclaw-memory-os/l2/{experience,value,standard}.md` ... L3 (native): managed block in `MEMORY.md`
Recommendation

Only capture durable information you actually want remembered, and periodically review the L2/L3 summaries and MEMORY.md managed block.

What this means

Future agents working in the workspace may follow the added memory-maintenance guidance.

Why it was flagged

The install flow intentionally persists agent-facing memory and maintenance instructions in workspace files.

Skill content
Run once to inject memory capability additively and generate migration + schedule artifacts ... Write only managed blocks in `AGENTS.md`, `MEMORY.md`, and optional `HEARTBEAT.md`.
Recommendation

Review AGENTS.md, MEMORY.md, optional HEARTBEAT.md, and the generated cron template before relying on them; remove the managed blocks if you do not want persistent behavior.

What this means

Approved or accidentally enabled cleanup actions can change how stored memory is summarized and rolled up.

Why it was flagged

The cleanup tool can mutate memory-entry statuses when actions have been enabled in the plan.

Skill content
enabled_actions = [action for action in plan.get("actions", []) if action.get("enabled")] ... entry.status = desired_status ... write_modified_files(file_lines)
Recommendation

Inspect the monthly cleanup plan JSON before enabling actions, use --dry-run first, and keep backups or version control for memory files.

What this means

If you enable the cron lines from an unusual or attacker-controlled path, the shell could interpret the path unexpectedly.

Why it was flagged

The generated cron template interpolates the workspace path into shell commands without quoting. It is not installed automatically, but the template could be unsafe if copied from a path containing shell metacharacters.

Skill content
f"{daily_minute} */2 * * * cd {workspace} && python3 skills/better-memory/scripts/run_daily_review.py --workspace . --threshold-only"
Recommendation

Inspect and shell-quote the workspace path before installing generated cron entries, or only use the template from a simple trusted path.