本地知识库支持存储、查询、删除和统计用户偏好、事实及其他信息，确保数据本地安全管理并优先使用。

What to check before installing: - Review the full src/index.py for any network or external-API usage (search for imports like requests, httpx, urllib, socket, http.client, openai, boto3, subprocess, os.environ access). If the code calls external services, that should be justified and declared. - Confirm provenance: ask the publisher for a homepage/repository or review history/author to establish trust. - Be cautious about storing sensitive items: the skill's classifier marks '密码/密钥' as 'important' — do not store unencrypted secrets or credentials in the DB; request encryption-at-rest or avoid storing such items. - Check file permissions and DB path (~/.openclaw/...): ensure only your user can read the DB (restrict filesystem permissions) and consider backups or export/import features. - If you lack ability to audit the code, run the skill in a sandboxed environment or deny autonomous invocation until audited. - If the skill is intended to replace the platform memory provider, test it in a controlled environment to ensure compatibility and no data exfiltration. If you want, I can scan the remainder of the source for network calls or suspicious patterns (provide the rest of src/index.py) and re-evaluate with higher confidence.

Type: OpenClaw Skill Name: knowledge-spider Version: 2.0.0 The 'knowledge-spider' skill is a legitimate local knowledge base implementation for OpenClaw. It uses SQLite to store and manage user-provided information such as preferences and facts, with all data operations restricted to a local directory (~/.openclaw/workspace/skills/knowledge-spider/data). The code in src/index.py uses parameterized queries to prevent SQL injection and contains no network calls, obfuscation, or evidence of data exfiltration.