mr-r0b0t - r0b0tlab ORG

Deploy concurrent Hermes Agent workers on unified-memory GPUs (GB10, DGX Spark) for maximum total tok/s. Profile-isolated, kanban-coordinated, crash-recovering.

macOS Hermes skill for sharing one canonical writable profile across Buzz and ACP surfaces

Filesystem-first LLM-Wiki + Obsidian + Hermes Agent memory system. Markdown source of truth, SQLite FTS5 search, secret scanning, tier-based memory. Built for local LLM setups.

DeepSeek-V4-Flash-DSpark optimized vLLM 0.26.0 SM121 dual-GB10 evidence (NVFP4 KV, B12X, DSpark K6)

DeepSeek-V4-Flash native Blackwell FP8 benchmark on dual DGX Spark GB10 (TP=2, MTP, RoCE). c=1 38.4 t/s, c=16 144.6 t/s aggregate.

GB10 NVFP4 native MTP reproducibility pack for Qwen3.6-35B-A3B

MiniMax M2.7 NVFP4 dual-GB10 Blackwell benchmark: vLLM FlashInfer-CUTLASS, public data, HTML canvas report, and Docker runtime.

No description.

Specification for a reproducible, provenance-bound multi-lane LLM benchmark suite.

Native SM121 vLLM runtime and reproducible qualification for Poolside Laguna S 2.1 NVFP4

Hermes Agent + Ansible GB10 cluster orchestration skill and runnable operator repo

NVIDIA Qwen3.6-27B NVFP4 on SM121 (GB10) — vLLM v0.24.0 with native NVFP4 KV cache via FlashInfer FA2 JIT. 67% more KV capacity than FP8.

Production-grade Hermes Agent skill for tldraw 5.2.5+ development, migration, sync, automation, and evaluation

Native SM121 vLLM deployment and benchmark evidence for NVIDIA Qwen3.6-35B-A3B-NVFP4 on one GB10

SM121-native Qwen3.6-35B-A3B NVFP4 Fast serving on NVIDIA GB10 with vLLM, FlashInfer B12X, FP8 KV, and MTP

Gemma-4-26B-A4B-NVFP4 native CUTLASS profile on GB10 / Blackwell

NVIDIA Nemotron 3.5 Lightning SM121 (GB10) reproducibility suite: NVFP4 MTP K=1 runtime, r0b0bench evidence, 1M-window NIAH probe

Step 3.7 Flash NVFP4 SM121-verified TP=2 on dual GB10 | vLLM + Ray

GB10 NVFP4 reproducibility pack for Gemma-4-31B-IT

Optimized SM121 vLLM container and benchmark report for nvidia/diffusiongemma-26B-A4B-it-NVFP4

NVFP4 quantized Nex-N2-mini (Qwen3.5-MoE-35B) — vLLM serving container for Blackwell SM121

GB10 NVFP4 native MTP reproducibility pack for Nemotron-3-Super-120B-A12B

NVFP4 FastContext weights, serve configs, and with-vs-without demo (standalone)

Agent Skills compliant skill for deploying ComfyUI on NVIDIA GB10 / DGX Spark with Hermes Agent integration. Optimized for sm_121 Blackwell, ARM64, 128GB unified memory.

Hy3-295B W4A4 NVFP4 v3 dual-GB10 benchmark and reproducibility evidence

FastContext-Hermes: Async subagent repo exploration on NVIDIA GB10

VibeThinker-3B NVFP4: SM121-optimized container for 3B reasoning model (2.60x throughput)

Sanitized GB10 AR benchmark results for unsloth/Qwen3.6-35B-A3B-NVFP4

SGLang SM121 reproducibility suite for NVIDIA Nemotron 3.5 Lightning NVFP4

No description.

Native SGLang serving for Inkling-Small NVFP4 on two NVIDIA GB10/SM121 nodes

SM121 vLLM 0.25 runtime for XYZ-Aquila-mini NVFP4

One dependable OpenAI-compatible endpoint for multiple explicit model pools

Qwen3.8-Max launch-day dossier — a self-portrait HTML artifact built by the model it documents. Official Qwen sources only.

The agent that grows with you

Hermes Agent model-provider plugin for Alibaba Cloud Token Plan (Global + China)

Collection of step-by-step playbooks for setting up AI/ML workloads on NVIDIA DGX Spark devices with Blackwell architecture.

Reachy Mini's SDK

No description.

vLLM v0.25.0 source-built for CUDA 13.0, ARM64, and NVIDIA GB10 SM121

Laguna XS 2.1 NVFP4 (33B MoE, 3B active) on GB10 SM121 — vLLM v0.24.0 native NVFP4 serving + DFlash Phase 2

SM121-native vLLM container for Agents-A1 NVFP4

Agent Skill for building OpenHome Abilities

Laguna M.1 NVFP4 on SM121 — dual GB10 TP=2 vLLM serve, concurrency benchmarks, GSM8K, Hermes agent micros

DeepSeek-V4-Flash NVFP4 deployment on dual GB10 (SM121) — vLLM + Ray multi-node

AgentSkills.io compliant Hermes skill for AgentMail — give AI agents their own email inbox