Download and cache model from HuggingFace (Ollama-like UX)
Usage: apr pull [OPTIONS] <MODEL>
Arguments:
<MODEL> Model reference (alias, hf:// URI, or org/repo)
Options:
--force Force re-download even if cached
--json Output as JSON
-v, --verbose Verbose output
-q, --quiet Quiet mode (errors only)
--offline Disable network access (Sovereign AI compliance, Section 9)
--skip-contract Skip tensor contract validation (PMAT-237: use with diagnostic tooling)
-h, --help Print help
-V, --version Print version