← Back to Articles & Artefacts
artefactssouth

INDEX — OpenClaw Mac Mini Infrastructure Research

IAIP Research
rch-tech-jgwill-claws-infrastructure

INDEX — OpenClaw Mac Mini Infrastructure Research

Date: April 15, 2026
For: Guillaume Descoteaux-Isabelle (jgwill)
Research Method: 3-cycle multi-agent orchestration (10 claude-opus-4.6 sub-agents)


Final Results

#DocumentDescription
1RESULT-01-openclaw-plugins-local-ai.mdOpenClaw plugin ecosystem — which plugins to install for local AI workflows (Ollama, HuggingFace, Copilot, Google, Perplexity), multi-provider routing, Hermes compatibility
2RESULT-02-mac-mini-inference-scenarios.mdMac Mini hardware for inference — Scenario A (minimal, ~$999) vs Scenario B (evolved, ~$2,399+), model memory requirements, benchmarks, pricing
3RESULT-03-mac-mini-training-scenarios.mdMac Mini hardware for training — MLX/MLX-LM frameworks, QMD model fine-tuning, LoRA persona adapters, data sovereignty protocols, automated weekend pipelines
4RESULT-04-copilot-google-plugin.mdGitHub Copilot + Google Plugin deep dive — model catalog, premium request quotas, Google multi-capability contracts, cost analysis, cloud+local routing

Orchestration Documentation

DocumentDescription
AGENTS.mdFull orchestration report — what agents ran, what they found, corrections applied, lessons learned

Research Pipeline (Raw Artefacts)

Cycle 0 — Initial Research (5 agents)

AgentFileFocus
A00-initial-research/agent-a-plugins.mdOpenClaw plugin ecosystem
B00-initial-research/agent-b-mac-inference.mdMac Mini M4 inference specs
C00-initial-research/agent-c-mac-training.mdApple Silicon training capabilities
D00-initial-research/agent-d-qmd-models.mdQMD HuggingFace models analysis
E00-initial-research/agent-e-copilot-google.mdCopilot + Google plugin

Cycle 1 — Review (2 agents)

ReviewerFileCovered
Track 101-review/review-track1-plugins-inference-copilot.mdAgents A, B, E
Track 201-review/review-track2-training-qmd.mdAgents C, D

Cycle 2 — Revision (3 agents)

ReviserOutputCorrections Applied
1RESULT-01 + RESULT-0415 items from Track 1 review
2RESULT-02Pricing, specs, benchmarks from Track 1
3RESULT-032 BLOCKING issues + 12 items from Track 2

PDE Decomposition

Quick Answers

What Mac Mini should I buy for inference?
→ See RESULT-02. Minimal: M4 24GB ($999). Large models: M4 Pro 64GB (~$2,399).

What Mac Mini should I buy for training?
→ See RESULT-03. Sweet spot: M4 Pro 48GB ($1,799). Max capability: Mac Studio M4 Max 128GB ($3,699).

Which OpenClaw plugins do I need?
→ See RESULT-01. Essential: ollama-provider. Useful: copilot-provider, google-plugin. Skip: minimax, moonshot.

Can I fine-tune QMD's models locally?
→ See RESULT-03. Yes, with caveats. Query expansion model has existing pipeline (CUDA-dependent). Embedding model fine-tuning is feasible but GGUF conversion is unverified.

What does the Google plugin do?
→ See RESULT-04. Multi-capability: Gemini LLM, image gen, video gen, web search grounding. Requires separate API key.