Long Term Memory 长期记忆

长期记忆管理系统 - 帮助AI和用户管理、存储、检索长期记忆。支持记忆分类、标签管理、重要性评分、自动压缩、跨会话记忆保持。适用于需要长期追踪信息、建立知识库、维护历史上下文的场景。

This skill mostly implements what it says (storing, searching, compressing memory files) but has several red flags you should resolve before installing or running it: 1) Do not trust the hardcoded BILLING_API_KEY in payment.py — it is sensitive and could indicate accidental key leakage or malicious use. Ask the publisher to remove the embedded key and require a configurable secret (and to document which env vars are required). 2) Clarify how billing is enforced: which scripts call payment.py, which environment variables are required (SKILLPAY_API_KEY vs SKILLPAY_USER_ID), and provide a clear, documented payment flow. 3) Verify missing referenced scripts (memory_organizer.py, memory_sync.py) or remove references to them. 4) Be aware the scripts read/write files under /root/.openclaw/workspace by default — run in a sandbox or set base_dir to a safe path and audit file contents before giving the skill access to sensitive data. If you cannot get satisfactory answers from the skill author, avoid installing or running it, especially on systems with sensitive data or credentials.

Type: OpenClaw Skill Name: shenmeng-long-term-memory Version: 2025.4.15 The skill implements a mandatory pay-per-use monetization system (SkillPay) that requires network calls to an external endpoint (skillpay.me) and uses a hardcoded API key in 'payment.py'. While the core memory management logic in 'memory_store.py', 'memory_search.py', and 'memory_compressor.py' appears functional for its stated purpose, the inclusion of a third-party billing gate for local file operations is highly unusual and poses privacy and security risks. No direct evidence of malicious data exfiltration or system compromise was found, but the non-standard payment requirement and hardcoded credentials warrant a suspicious classification.