skills/xberg-io/plugins/liter-llm/Gen Agent Trust Hub

liter-llm

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and installs multiple components from official vendor-owned sources.
  • Fetches binaries and source code from the author's GitHub repository: https://github.com/xberg-io/liter-llm.
  • Downloads Docker images from the GitHub Container Registry: ghcr.io/xberg-io/liter-llm.
  • Installs language-specific packages from standard registries (NPM, PyPI, Rubygems, etc.) under the @xberg-io organization or matching package names.
  • [COMMAND_EXECUTION]: The documentation provides instructions for executing CLI commands to manage LLM interactions and proxy services.
  • Includes commands for package managers (npm, pip, cargo, brew) and runtime execution (npx, uvx).
  • Mentions a local launch script scripts/mcp-launch.sh for auto-registering the Model Context Protocol (MCP) server.
  • [CREDENTIALS_UNSAFE]: The skill handles sensitive API keys for multiple LLM providers.
  • Demonstrates safe practice by reading keys from environment variables (e.g., os.environ["OPENAI_API_KEY"]) rather than hardcoding them.
  • Explicitly mentions that keys are wrapped in SecretString types to prevent accidental logging or serialization.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 10:22 PM
Security Audit — agent-trust-hub — liter-llm