Back to skill

Security audit

Claude Clarity

Security checks for vulnerabilities and agentic risk

Overview

This is a broad local cognitive-memory skill that is intrusive by design, but its high-impact behavior is mostly disclosed and purpose-aligned.

Install only with the expectation that it will alter local agent configuration and add a persistent local memory/MCP layer. Review the settings.json and CLAUDE.md changes, keep network and code-execution related options disabled unless intentionally needed, and treat the strong persona behavior as part of the skill’s design.

SkillSpector could not complete.