unified-ai-gateway
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides detailed shell instructions for the agent to manage Docker operations on the host system. These include pulling images, inspecting container metadata, and exporting filesystems to temporary directories for static analysis.
- [COMMAND_EXECUTION]: It utilizes shell-based search tools (find, grep, sha256sum) to audit the extracted container contents for potential risks like native binaries, hardcoded credentials, and lifecycle hooks before the tool is registered.
- [EXTERNAL_DOWNLOADS]: The skill fetches a Docker image from the GitHub Container Registry (
ghcr.io), which is a well-known and reputable service. The download is cryptographically verified using a specific SHA-256 digest to ensure supply chain integrity. - [REMOTE_CODE_EXECUTION]: The skill facilitates the registration of a local MCP server that executes via a Docker container. Security risks associated with remote execution are mitigated by enforcing strict runtime constraints: no network access (
--network none), dropping all Linux capabilities (--cap-drop ALL), and preventing privilege escalation (--security-opt no-new-privileges). - [SAFE]: The skill demonstrates high security maturity by explicitly documenting risks—such as the container running as root—and requiring manual user approval for filesystem changes and tool registration.
Audit Metadata