v1.0.0

Agent Memory Patterns

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 5:16 AM.

Analysis

This is a coherent instruction-only memory guide, but it describes persistent agent memories that may retain user, project, and external-content information.

GuidanceThis skill appears benign and purpose-aligned as a memory-pattern guide. Before installing or using it, be aware that its examples are designed to preserve agent memory over time; do not store secrets in those files, periodically review or delete stale memories, and verify external content before it influences future agent behavior.

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.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Rogue Agents

SeverityInfoConfidenceMediumStatusNote

SKILL.md

heartbeat-state.json # ハートビート状態 ... queued-messages.json # メッセージキュー

The architecture includes persistent agent state and a message queue, which fits the stated persistent-agent memory purpose but is still behavior users should understand before enabling.

User impactThe agent may be designed to maintain state across sessions, which can affect future behavior if the stored state is inaccurate or stale.

RecommendationKeep persistent state transparent, user-reviewable, and easy to clear or disable.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Memory and Context Poisoning

SeverityLowConfidenceHighStatusNote

SKILL.md

MEMORY.md # 長期記憶（手動キュレーション） ... memory/YYYY-MM-DD.md # 日次ログ ... pending-memories.md # 外部コンテンツ段階処理

The skill intentionally stores long-term memory, daily logs, and queued external content for later reuse, which is purpose-aligned but creates persistent context that users should review and curate.

User impactInformation from conversations, projects, or external sources may be kept and reused in later agent sessions.

RecommendationUse clear retention and review practices for memory files, avoid storing secrets, and manually validate external content before promoting it into long-term memory.