What problem does it solve?
Unvetted MCP server registrations in settings.json can introduce critical security risks including privilege escalation, malicious URL access, shell injection vulnerabilities, and unauthorized tool exposure before the runtime loads the configuration. This skill eliminates those risks by enforcing strict validation gates at registration time.
Core Features & Use Cases
- Allowlist and URL validation: Checks all MCP server entries against the kit's allowlist, enforces HTTPS scheme rules, and blocks non-localhost HTTP URLs to prevent unauthorized remote access.
- Command and tool safety auditing: Validates local command-based MCP entries for valid package sources, rejects shell metacharacter injection risks, and blocks wildcard tool exposure to enforce least-privilege access.
- Policy and threat compliance: Enforces tier classification limits (max 5 context-tier servers), emits STRIDE threat delta reports for new registrations, and supports drift checks between policy tables and runtime configurations.
Use case: A developer adding a new third-party MCP server to their project's settings.json can run this skill to confirm the entry complies with all security policies before committing the change, avoiding runtime security incidents.
Quick Start
Use the mcp-permission-validate skill to audit all MCP server entries in your project's settings.json file for security and policy compliance.
