Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add LuuOW/meridian-mcp --skill mlopsgit clone --depth 1 https://github.com/LuuOW/meridian-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/luuow/meridian-mcp/mlops)<a href="https://agentmods.dev/skills/luuow/meridian-mcp/mlops"><img src="https://agentmods.dev/badge/skills/luuow/meridian-mcp/mlops.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00072 | $0.01914 |
| Opus 5 | $0.00036 | $0.00957 |
| Sonnet 5 | $0.00014 | $0.00383 |
| Haiku 4.5 | $0.00007 | $0.00191 |
Grade A, and why
mlops scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MLOps
MLOps is the discipline of applying DevOps principles to the full machine-learning lifecycle: data ingestion, feature engineering, training, evaluation, serving, monitoring, and retraining. This skill covers the infrastructure layer — the plumbing that keeps models reliable in production — not model architecture or research. Expect opinionated guidance on versioning, deployment strategies, observability, and the operational patterns that separate proof-of-concept from production systems.
Core Concepts
Model Serving
vLLM is the dominant LLM inference server. Key knobs: --tensor-parallel-size (split across GPUs), --gpu-memory-utilization (default 0.9, lower if OOM during prefill), --max-model-len (cap KV-cache footprint). PagedAttention means KV cache is allocated in non-contiguous blocks — profiling with --disable-log-stats=false exposes scheduler queue depth. For multi-LoRA serving use --enable-lora with --max-loras and load adapters via the /v1/load_lora_adapter endpoint without restarting.
Triton Inference Server uses an ensemble model type to chain pre/post-processing with inference in a single request. Each model directory needs a config.pbtxt. Dynamic batching: set max_queue_delay_microseconds and preferred_batch_size. Use the ONNX backend for portability; TensorRT backend for latency-critical paths. Monitor nv_inference_queue_duration_us and nv_inference_exec_count in Prometheus.
TorchServe handler pattern: subclass BaseHandler, override preprocess, inference, postprocess. Register with torch-model-archiver --model-name foo --version 1.0 --serialized-file model.pt --handler handler.py. Management API on :8081, inference on :8080. Scale workers per model: PUT /models/foo?min_worker=2&max_worker=8.
BentoML is useful when you want a Python-native abstraction over multiple backends. @bentoml.service + @bentoml.api decorators; deploy to BentoCloud or export as OCI image.
Feature Stores
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 77 lines · 72 tokens per session scan A 906dadac34f2
mlops is a skill published in the GitHub repository LuuOW/meridian-mcp (0 stars, last pushed 3d ago), licensed MIT. It adds 72 tokens to every session and 1,914 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
mlflow
Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform.
weights-and-biases
Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B - collaborative MLOps platform.
weights-and-biases
Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B - collaborative MLOps platform.
mlflow
Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform.
weights-and-biases
Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B - collaborative MLOps platform.
mlflow
Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow - framework-agnostic ML lifecycle platform.