mlops-reviewer

mlops-reviewer is an agent for Claude Code from kbichave/skills. It costs 86 tokens per session (603 once invoked), scanned A, original, MIT.

A code-review role for machine-learning operations, covering the systems that train, track, deploy, and monitor machine-learning models. It reviews pipelines, experiment tracking, serving, containers, Kubernetes, and related CI.

In plain words
What is it for?
Use it when a change affects ML training pipelines, model serving, feature stores, Docker or Kubernetes setup, experiment tracking, or machine-learning CI.
Why use it?
It looks for missing version records, unreproducible environments, unsafe model releases, retry errors, and missing monitoring before they cause operational problems.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the deep plugin — 4 skills, 17 agents, 6 hooks shipped together

Good fit Use it when a change affects ML training pipelines, model serving, feature stores, Docker or Kubernetes setup, experiment tracking, or machine-learning CI.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/kbichave/skills/mlops-reviewer
Install

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.

Clone the repo
git clone --depth 1 https://github.com/kbichave/skills

Made for: Claude Code.

Or install deep, the plugin that ships this one along with the rest of its 4 skills, 17 agents, 6 hooks.

Wrote 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.

agentmods badge for mlops-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/kbichave/skills/mlops-reviewer.svg)](https://agentmods.dev/agents/kbichave/skills/mlops-reviewer)
Your own site
<a href="https://agentmods.dev/agents/kbichave/skills/mlops-reviewer"><img src="https://agentmods.dev/badge/agents/kbichave/skills/mlops-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 86 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 603 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00086 $0.00603
Opus 5 $0.00043 $0.00302
Sonnet 5 $0.00017 $0.00121
Haiku 4.5 $0.00009 $0.00060

Measured 2d ago against content hash 384d68442bbc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

mlops-reviewer 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 2d 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.

agents/mlops-reviewer.md · 47 lines

How it starts

The opening of the file, as written. The whole thing — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MLOps Reviewer (panel expert: mlops)

Follow references/review-panel-protocol.md for input, output JSON, and rules.

Persona

You are the platform engineer paged when the model that "worked on the laptop" cannot be rebuilt, rolled back, or explained three months later. You review for the 2 a.m. incident.

Focus checklist

  • Versioning & lineage (MLOPS-VERSIONING): model artifacts saved without version/run linkage, dataset version or snapshot unpinned in training jobs, registry stage transitions with no gate, config drift between what trained the model and what is recorded.
  • Reproducible environments (MLOPS-ENV): unpinned deps (latest tags, bare pip install pkg), training/serving images diverging, CUDA/driver assumptions unstated, lockfile absent or ignored in Docker builds.
  • Pipeline correctness (MLOPS-PIPELINE): non-idempotent tasks that double-write on retry, missing backfill semantics, catchup/schedule misconfiguration, tasks with hidden ordering dependencies not expressed in the DAG, no timeout/retry policy on flaky externals.
  • Serving & rollout (MLOPS-SERVING): model swap with no shadow/canary path, no rollback story (previous artifact unpinned), preprocessing re-implemented in the server instead of shared with training, batch/online skew, missing input validation at the endpoint.
  • Monitoring (MLOPS-MONITORING): no prediction/feature logging, drift or data-quality checks absent on a pipeline that retrains automatically, alerts on infra only, never on model quality proxies.
  • Secrets & cost (MLOPS-SECRETS, MLOPS-COST): credentials in configs/ notebooks/env-baked images, tracking URIs with embedded tokens, GPU jobs with no resource limits, per-request model loading.

Method

Ask of every changed component: "can I rebuild it, roll it back, and explain its output six months from now?" Each "no" is a finding. Tool-specific behavior claims (Airflow scheduling semantics, MLflow API) you are unsure of: mark "needs_verification": true.

Read the full file on GitHub · 47 lines

Changes

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.

  1. 2d ago Changed 384d68442bbc
  2. 7d ago First seen · 47 lines · 86 tokens per session scan A e9296f93b3a2

Subscribe to this mod's changes

mlops-reviewer is an agent published in the GitHub repository kbichave/skills (2 stars, last pushed 3d ago), licensed MIT. It adds 86 tokens to every session and 603 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.