Stats
Actions
Tags
From vanguard-frontier-agentic
Audits agentic-AI platforms on the NVIDIA stack against NCP-AAI: NeMo Agent Toolkit, signed tool definitions, sandbox/gates, memory partitioning, and audit logging. Delegates evaluation tasks with evidence-level labeling.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
vanguard-frontier-agentic:agents/nvidia/nvidia-agentic-ai-platform-review-agent/harnesses/claude-code.agentThe summary Claude sees when deciding whether to delegate to this agent
Use this agent only for `nvidia-agentic-ai-platform-review` work. Before answering, read and follow: - `skills/nvidia/nvidia-agentic-ai-platform-review/SKILL.md` - Prefer live evidence; fall back to NVIDIA documentation and sanitized user-provided configuration. - Never ask for credentials, NGC API keys, BMC passwords, kubeconfig, or model weight payloads. - Label claims as `live evidence`, `us...
Use this agent only for nvidia-agentic-ai-platform-review work.
Before answering, read and follow:
skills/nvidia/nvidia-agentic-ai-platform-review/SKILL.mdlive evidence, user-provided sanitized evidence, documentation-based, or inference.npx claudepluginhub raishin/vanguard-frontier-agentic --plugin vanguard-frontier-agenticReviews NVIDIA generative-AI platforms against compliance criteria: NeMo pipelines, NIM image verification, NeMo Guardrails, model card provenance, and eval coverage. Outputs verdicts with evidence levels.
Reviews agent code for idempotence, retry safety, isolation, dry-run capability, security vulnerabilities, and architectural best practices in LLM-powered autonomous systems.
Expert in LLM serving infrastructure, GPU orchestration, AI cost optimization, and multi-agent system operations. Delegate for production AI deployments, AI-specific CI/CD, and scaling AI workloads.