Structured Vector Memory (SVM)
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill matches its memory-management purpose, but it warrants review because it sets up scheduled background jobs that read past conversations and store or reuse them as persistent memory.
Before installing, decide whether you are comfortable with automatic memory jobs reviewing past conversations and persisting summaries. If you proceed, review the cron entries, restrict what sessions can be captured, require approval before modifying SYSTEM_GUIDE.md, and audit the referenced memory-lancedb-pro plugin and API-key handling.
Findings (5)
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.
Sensitive or incorrect details from past conversations may be stored and later reused by the agent.
The scheduled task is designed to review all prior-day conversation sessions and write persistent summaries, creating broad stored memory from private chats.
- [ ] 回顧昨天（$YESTERDAY）所有對話 session - [ ] 生成結構化摘要，寫入 second-brain/summaries/$YESTERDAY.md
Install only if you want broad automatic memory. Add retention limits, exclusions for sensitive sessions, and a review step before storing summaries.
Memory tasks may run in the background on a schedule rather than only when you ask for them.
The skill instructs the user to install recurring cron jobs that keep running after setup.
0 10,13,16,19,22 * * * ~/.openclaw/shared/micro-sync.sh >> ~/.openclaw/logs/micro-sync.log 2>&1 0 1 * * * ~/.openclaw/shared/daily-wrapup.sh >> ~/.openclaw/logs/daily-wrapup.log 2>&1 0 3 * * 0 ~/.openclaw/shared/weekly-compound.sh >> ~/.openclaw/logs/weekly-compound.log 2>&1
Review the crontab entries before installing, monitor the logs, and remove the cron jobs if you do not want ongoing automation.
Future agent behavior could change automatically if the task updates core guidance files without your review.
A scheduled memory-maintenance task can instruct the agent to modify SYSTEM_GUIDE.md, which may affect future agent behavior beyond simple memory summarization.
6. **檢查本週有無架構性變更（如新增 agent/workflow），如有，更新 SYSTEM_GUIDE.md**
Require a visible diff and explicit user approval before modifying SYSTEM_GUIDE.md or other agent-control files.
You may need to provide an external API key for the memory stack to work.
The skill expects a provider API key for embeddings and reranking; this is purpose-aligned, but it is a credential users should scope and protect.
- Jina API Key（embedding + rerank）
Use a least-privilege key, store it only in the expected plugin configuration, and review the memory-lancedb-pro plugin separately.
The safety of the full setup also depends on the separate memory-lancedb-pro plugin.
The skill depends on another plugin that is not included in the reviewed artifacts, so this review cannot verify that plugin's behavior.
- memory-lancedb-pro 插件已安装（`clawhub install memory-lancedb-pro`）
Review and trust the referenced plugin before enabling autoCapture or autoRecall.