Bernstein is an open-source governance layer for AI agents that applies written policies, schedules agent work deterministically, controls approvals, and records verifiable execution history. It is used to coordinate CLI coding agents and other agent workloads, including parallel runs and offline review, while the catalogue entries provide its agent workflows and controls.
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 agentmods add skills/sipyourdrink-ltd/bernstein/ml-engineernpx skills add sipyourdrink-ltd/bernstein --skill ml-engineergit clone --depth 1 https://github.com/sipyourdrink-ltd/bernsteinWrote 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/sipyourdrink-ltd/bernstein/ml-engineer)<a href="https://agentmods.dev/skills/sipyourdrink-ltd/bernstein/ml-engineer"><img src="https://agentmods.dev/badge/skills/sipyourdrink-ltd/bernstein/ml-engineer.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 | $0.00014 | $0.00323 |
| Opus 5 | $0.00007 | $0.00161 |
| Sonnet 5 | $0.00003 | $0.00065 |
| Haiku 4.5 | $0.00001 | $0.00032 |
Grade A, and why
ml-engineer 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 5d 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.
What it actually says
ML Engineering Skill
You are an ML engineer. Build, train, evaluate, and deploy machine learning models and inference pipelines.
Specialization
- Model training and fine-tuning (PyTorch, Transformers)
- Embedding models and vector representations
- RAG pipelines and retrieval-augmented generation
- Inference optimization (quantization, batching, caching)
- Evaluation metrics and experiment tracking
- Data preprocessing and feature engineering
Work style
- Read the task description and existing pipeline code before writing.
- Start with a clear hypothesis and success metric for every change.
- Write deterministic tests for data transforms and scoring logic.
- Keep model configuration separate from training/inference code.
- Log metrics, parameters, and artifacts for reproducibility.
Rules
- Only modify files listed in your task's
owned_files. - Run tests before marking complete:
uv run python scripts/run_tests.py -x. - Never commit model weights or large data files to git.
- Document any new dependencies in
pyproject.toml.
Call load_skill(name="ml-engineer", reference="evaluation.md") for
metric guidance, or reference="reproducibility.md" for experiment
tracking rules.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 5d ago First seen · 47 lines · 14 tokens per session scan A 2de54ac93325
ml-engineer is a skill published in the GitHub repository sipyourdrink-ltd/bernstein (1,099 stars, last pushed today), licensed Apache-2.0. It adds 14 tokens to every session and 323 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
tao-validate-recipe-transfer
Port a published computer vision paper's official code and training recipe onto a customer's own dataset, or diagnose why such a transfer produced bad numbers. Use this whenever someone wants to reproduce a CV paper, run a paper's repo on their own images, fine-tune a published…
deep-learning
Use when training or debugging a neural net in PyTorch — the forward/loss/backward/step loop and its silent bugs, mixed precision (AMP), AdamW/LR schedules, DDP/FSDP/ZeRO, checkpoints and seeds. NOT LoRA/QLoRA on a pretrained LLM (that is finetuning), NOT tabular sklearn/XGBoost (that is machine-learning), NOT…
add-pipeline-stage
Add a single new stage to an existing Calkit pipeline. Use when the user invokes /calkit:add-pipeline-stage or asks to add a script, notebook, or command to the pipeline.
bio-workbench
生信分析工作台（dsh-science-workbench）的可复现分析约定。当用户要创建或维护一个可复现的生信分析项目、用 bioruncell 跑分析并出图、登记产物 provenance、对图做结构化反馈并重画（反馈→改代码→重跑→派生新版本）时使用此 skill。触发词：生信项目、分析工作台、bioruncell、可复现出图、manifest、cell 契约、反馈重画。.
prepare-data
Profile and prepare raw source files before modeling or loading them into Cruxible; validate keys, grain, joins, and transformation needs, then produce a concrete readiness report.
text-provenance
Use when you need to identify the likely source of a text passage, attribute text to documents in a RAG system, detect plagiarism, or match contract clauses to their origin.