skills/1broseidon/skills/folio/Gen Agent Trust Hub

folio

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes internal scripts (scripts/serve.js and scripts/share.js) that execute system commands to facilitate the local development lifecycle. serve.js implements a static file server using the Node.js http module with built-in path traversal guards, while share.js utilizes standard ssh or ngrok commands to establish ephemeral tunnels to the local host.
  • [EXTERNAL_DOWNLOADS]: The skill's host profiles and documentation instruct the agent to use npx for executing deployment tools such as surge, gh, and wrangler. The skill also manages external dependencies (e.g., D3.js, Chart.js) from well-known content delivery networks like jsdelivr.net through a mandatory manifest ledger system.
  • [SAFE]: The skill enforces multiple security-focused disciplines, including a 'static-only' rule that prohibits generated artifacts from communicating with external APIs or WebSockets. It explicitly forbids the storage of secrets in source files and requires labeling mock data. The server implementation in scripts/serve.js includes specific checks to prevent directory traversal attacks by validating that resolved file paths remain within the intended site root.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 08:31 PM
Security Audit — agent-trust-hub — folio