fabius-doctrinalisted

fabius's AI/ML-engineering layer — train, serve, evaluate, and operate machine-learning and LLM systems as production software. The model lifecycle: dataset → train / fine-tune → EVALUATE → serve / infer → monitor. It owns model serving and inference (vLLM-class, OpenAI-compatible endpoints, batching, quantization), MLOps and experiment tracking (MLflow-class), and rigorous model/LLM evaluation (eval harnesses, blind judges, regression gates). It is NOT agent orchestration — that's fabius-cohors (doctrina owns the MODEL an agent calls, not the agent). Use when the task is to train / fine-tune a model, serve or deploy a model, evaluate a model or prompt, track ML experiments, set up an inference endpoint, build a RAG/LLM-app's model tier, or when the user says "serve this model", "fine-tune", "eval my prompts", "track these runs", "MLOps", "why is inference slow", or names vLLM / MLflow / an eval harness.
shear559/fabius · ★ 2 · AI & Automation · score 64

Install: claude install-skill shear559/fabius

# Fabius Doctrina — train it, serve it, prove it, operate it *Doctrina* — the body of learning, the act of teaching something to learn. Machine **learning** is its literal subject: a model is taught from data, then it has to earn production. A model that scores well in a notebook and silently rots in production is the failure mode here — so the same Fabian stance holds, made specific to models: **scout the data and the metric, strike the smallest model that clears the bar, and prove it on a held-out reality before it serves a user.** ## 1. Doctrina and its neighbors — one concern, sharp edges - **doctrina** owns the **model as a built artifact**: train / fine-tune it, evaluate it, serve it, operate it. The risk is a model that looks right offline and fails on live inputs. - **fabius-scientia** owns the **scientific method over natural-science data** (hypotheses, omics/bio databases, reproducible analysis). A gene-expression study is scientia; serving a trained classifier is doctrina. Both share the reproducibility discipline; they don't share the concern. - **fabius-cohors** owns the **agent** (LLM tools, orchestration, output contracts). doctrina owns the **model the agent calls** — its serving, its evaluation, its cost. An agent that calls a model: cohors owns the agent loop, doctrina owns whether the model is served well an