lvtd-skills-router
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides Node.js scripts and shell commands to query a local registry file (
dist/registry.json). These commands are used for searching skill metadata (names, descriptions, tags) and do not interact with sensitive system files or external network resources. - [DYNAMIC_EXECUTION]: Instructions direct the agent to execute Node.js one-liners to process JSON data and route user requests. These scripts rely on standard Node.js built-ins like
node:fs/promisesto read local repository artifacts. - [METADATA_POISONING]: The skill includes logic to prevent manual catalog maintenance by relying on a generated
registry.json. This is a best practice that reduces the risk of stale or misleading metadata within the router skill itself.
Audit Metadata