llm-models

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and directs the user to fetch installation scripts and documentation from the inference-sh public GitHub repository.
  • [COMMAND_EXECUTION]: The instructions utilize the belt CLI tool to perform authentication and execute model inference requests. The tool usage is scoped within the allowed-tools frontmatter configuration.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides a surface for processing untrusted data by passing user-defined prompts to external language models.
  • Ingestion points: User input provided via the --input flag in shell commands and external input.json files.
  • Capability inventory: The skill uses CLI tools with network capabilities to send data to third-party AI providers (OpenRouter, Anthropic, Google, etc.).
  • Boundary markers: None identified in the instructional examples.
  • Sanitization: Content filtering is handled by the downstream model providers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:20 PM
Security Audit — agent-trust-hub — llm-models