local-llm-provider

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The setup instructions in the documentation describe installing Ollama via a piped script: curl -fsSL https://ollama.com/install.sh | sh. This is a standard installation pattern for the official Ollama service.
  • [EXTERNAL_DOWNLOADS]: The skill references external resources, including the Ollama installation script and the vllm package available on PyPI, for local inference capabilities.
  • [COMMAND_EXECUTION]: The skill provides CLI-based entry points for interacting with the local-llm-provider logic, allowing users to run queries and manage model listing through Node.js commands.
  • [PROMPT_INJECTION]: The skill processes untrusted user prompts by interpolating them into JSON payloads for LLM provider requests.
  • Ingestion points: The prompt variable in the complete method of provider.js and input via queryCLI.
  • Boundary markers: No delimiters or explicit safety instructions are used to wrap user input.
  • Capability inventory: The skill makes network requests to local and remote API endpoints.
  • Sanitization: No evidence of input validation or content sanitization before API transmission.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 05:11 AM