将外部内容视为不可信数据，使用前需验证，并在沙箱环境中隔离检查。
38 次安装1 星标
安全
LyraShield AI
试用Review LyraShield release-assurance evidence through an OAuth-first MCP connection.
它能做什么
Review LyraShield release-assurance evidence through an OAuth-first MCP connection.
技能文档
LyraShield AI review skill
Use the LyraShield MCP connection to inspect release-assurance evidence, summarize current issues, and explain pending approvals. Prefer read-only tools. Do not claim that a result is a security guarantee or that all Vibe Security 50 controls were verified.
The skill is a community ClawHub listing. It is not an official OpenClaw channel. Link to LyraShield dashboard evidence instead of recreating dashboard UI or copying report contents into prompts.
Write actions require OAuth lyrashield.write plus the exact-argument approval gate.
相关技能
Code quality audit guidance, security review, vulnerability identification patterns, and dependency risk assessment. AI-delivered service via clawtip verification.
Security and compliance auditing tool for AI agents. Scans code for vulnerabilities, checks GDPR/CCPA compliance, generates risk reports with remediation guidance.
在本地运行 77 项安全测试,为 AI 代理签发证书并与其他受信代理完成握手。
67 次安装
Create CompleteTech LLC security, safety, permissions, and production-readiness review artifacts for agentic development workflows, including risk intake, to...
11 次安装
Independent fail-closed second opinion before acting: allow/review/block a risky action, fact-check a claim, screen text for prompt injection, or flag PII/se...