agentic-gateway

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of command-line utilities such as curl, npx, node, jq, grep, and cut. These tools are used for prototyping API calls, bootstrapping wallet environments, and generating cryptographic signatures. This behavior is entirely consistent with a tool designed for developers and application integration.
  • [EXTERNAL_DOWNLOADS]: Users are instructed to install various npm packages including @alchemy/x402, viem, mppx, and @stripe/stripe-js. These are official vendor packages or established industry-standard libraries. No downloads from untrusted or anonymous sources were detected.
  • [DATA_EXFILTRATION]: Although the skill handles sensitive information (blockchain private keys), it contains specific 'Hard requirements' that forbid the agent from using its Read, Write, or Edit tools on files like wallet-key.txt or .env. It correctly advises using shell pipes to manage these keys, which prevents sensitive data from being captured in the agent's context or tool logs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 10:30 AM
Security Audit — agent-trust-hub — agentic-gateway