influencer-discovery

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions include curl commands and Python urllib calls to fetch API tokens and register OAuth clients via api.xpoz.ai and mcp.xpoz.ai domains.
  • [COMMAND_EXECUTION]: The skill uses subprocess.run to execute the mcporter CLI tool for local configuration. It passes an OAuth token retrieved from the vendor's API as an argument to this command.
  • [REMOTE_CODE_EXECUTION]: An automated scan flagged a pattern where data from https://mcp.xpoz.ai/oauth/token is processed and then used in a subprocess. Analysis confirms this is a standard OAuth token exchange used to configure the local environment, targeting the vendor's own infrastructure.
  • [SAFE]: All identified external resources, including the Python package xpoz, the NPM package @xpoz/xpoz, and the xpoz.ai subdomains, are owned and maintained by the skill's author, XPOZpublic.
  • [DATA_EXPOSURE]: The skill writes temporary OAuth state data to ~/.cache/xpoz-oauth/state.json. This is standard practice for state management during the authentication flow and does not target sensitive system files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 08:04 AM
Security Audit — agent-trust-hub — influencer-discovery