nemo-relay-get-started

Pass

Audited by Gen Agent Trust Hub on Oct 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill incorporates strong security boundaries, including the use of deterministic 'smoke prompts' that avoid sensitive resources and mandatory user verification for all configuration changes. Instructions also strictly prohibit the agent from printing or exposing API keys, tokens, or authentication files, advising the use of existing environment-based authentication.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests local project manifest files to configure the environment, representing an indirect injection surface.\n
  • Ingestion points: Reads project manifest files (package.json, pyproject.toml, Cargo.toml) and Relay configuration files.\n
  • Boundary markers: Explicitly requires the agent to preview all changes and obtain user confirmation before writing files or executing code.\n
  • Capability inventory: Modifies code, writes configuration, and executes shell commands via the nemo-relay CLI.\n
  • Sanitization: Relies on human-in-the-loop review and structured trial protocols to prevent execution of unauthorized instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 3, 2026, 05:19 PM