Skip to content
Back to search

probabl-skills

MarketplaceWorks with Claude CodeGitHub

Tabular Data Science Skills for guardrailing AI Agents

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/probabl-ai/skore-skills with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/probabl-ai/skore-skills'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/probabl-ai--skills/badge.svg)](/listings/probabl-ai--skills)

HTML

<a href="/listings/probabl-ai--skills"><img src="/api/listings/probabl-ai--skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

You can install the skills using the `skore` CLI that you can install from PyPI or from conda-forge and run the following command.

First install [skore-cli](https://github.com/probabl-ai/skore-cli):

# with pip
pip install skore-cli
# with uv
uv tool install skore-cli
# with pixi
pixi global install skore-cli

Then run the following command:

skore skills install

Install a smaller workflow pack by id when you do not need the full companion:

skore skills install setup  # workspace, environment, git, export
skore skills install data_analysis  # data exploration
skore skills install model  # frame, build, smoke-test, evaluate, audit
skore skills install loop   # triage, explore, model, review, backlog, export
skore skills install export  # notebooks and documentation site

`skore skills install ml-experimentation` remains the complete pack, containing every active skill. The default `install` / `install all` behavior is unchanged.

You can use `uvx` or `pixi exec` to install the `skore` CLI and directly run the command in an isolated environment:

uvx --from skore-cli skore skills install

or

pixi exec --spec skore-cli skore skills install

If you prefer `npx`, then you can use:

npx skills add probabl-ai/skore-skills

Alternative — Claude Code plugin marketplace

If you only use Claude Code and prefer the native plugin flow, this repo is also a [Claude Code plugin marketplace](https://docs.claude.com/en/docs/claude-code/plugin-marketplaces):

/plugin marketplace add probabl-ai/skore-skills
/plugin install probabl-skills@probabl-skills

`/plugin update` pulls new releases.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 33 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "$schema": "https://json.schemastore.org/claude-code-plugin-marketplace.json",
  3. "name": "probabl-skills",
  4. "owner": {
  5. "name": "Probabl",
  6. "url": "https://probabl.ai"
  7. },
  8. "metadata": {
  9. "description": "Skills for ML experimentation in Python, organized around skrub, scikit-learn, and skore."
  10. },
  11. "plugins": [
  12. {
  13. "name": "probabl-skills",
  14. "source": "./",
  15. "description": "Workflow skills for ML experimentation with skrub, scikit-learn, and skore: pipeline build / evaluate / test / audit, iteration loop, workspace + tooling.",
  16. "author": {
  17. "name": "Probabl",
  18. "url": "https://probabl.ai"
  19. },
  20. "homepage": "https://github.com/probabl-ai/skore-skills",
  21. "license": "BSD-3-Clause",
  22. "keywords": [
  23. "machine-learning",
  24. "data-science",
  25. "scikit-learn",
  26. "skrub",
  27. "skore",
  28. "ml-experimentation"
  29. ],
  30. "category": "data-science"
  31. }
  32. ]
  33. }
README.md · README · 184 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # Probabl Skills
  2. A set of skills to steer your AI-assisted machine learning experiments.
  3. The skills help you:
  4. - build your machine learning pipeline with core data science libraries
  5. (e.g. scikit-learn, skrub, skore, pandas, polars) while ensuring
  6. your agent follows correct methodologies
  7. - evaluate and store your results so you can easily audit and get insights from them
  8. - connect your agent to [Skore Hub](https://skore.probabl.ai/) to get a comprehensive view of
  9. your experiments and results
  10. - iterate on your next experiments from a Skore audit digest (a
  11. separate backlog turn) and from your own feedback
  12. - organize your workspace according to best practices for data science projects
  13. (e.g. cookiecutter template)
  14. Probabl skills let you focus on the science while AI agents handle the implementation,
  15. guided by two important ingredients: core data science libraries for maintainability and
  16. methodological best practices for running your machine learning experiments properly.
  17. In practice, from a prompt such as:
  18. ```text
  19. ╭────────────────────────────────────────────────────────────────────────╮
  20. │ > Given the context in the file `data/README.md` and the data located │
  21. │ in `data/`, let's build a first machine learning pipeline that will │
  22. │ serve as baseline for the next experiments that we are going to run │
  23. │ together. │
  24. ╰────────────────────────────────────────────────────────────────────────╯
  25. ```
  26. you can expect your agent to start experimenting with you. The skills work well with
  27. models such as Claude Opus and Sonnet and produce great results with smaller models such
  28. as Qwen 3.7 Flash or DeepSeek v4.1 Flash. Each skill declares `metadata.modelTier`
  29. (`small`, `medium`, or `big`). The eval maps those tiers to Qwen3.8 27B,
  30. DeepSeek v4.1 Flash, and Qwen3.8 Max. A skill may also declare `metadata.role`
  31. (`entry` or `helper`) when its caller continues after it loads. A helper runs on at
  32. most its caller's tier, so a skill never loads a helper of a higher tier;
  33. `tools/validate_catalog.py` enforces it.
  34. As for agent harnesses, we tested them with Claude Code, OpenCode, Cursor, and GitHub
  35. Copilot and found no significant difference in terms of skill invocation.
  36. ## Install
  37. The skills installer is the `skore` command, included in the [`skore`](https://github.com/probabl-ai/skore)
  38. package on PyPI and conda-forge.
  39. ```bash
  40. # with pip
  41. pip install skore
  42. # with uv
  43. uv tool install skore
  44. # with pixi
  45. pixi global install skore
  46. # with conda
  47. conda install conda-forge::skore
  48. ```
  49. `skore skills install` with no arguments opens a picker for the source, the
  50. skills, the agent, and the project or user directory. Without a terminal, that
  51. same command prints the catalog.
  52. Install a workflow pack by id when you do not need every skill:
  53. ```bash
  54. skore skills install setup # workspace, environment, git, export
  55. skore skills install data_analysis # data exploration
  56. skore skills install model # frame, build, smoke-test, evaluate, audit
  57. skore skills install loop # triage, explore, model, review, backlog, export
  58. skore skills install export # notebooks and documentation site
  59. ```
  60. `skore skills install ml-experimentation` and `skore skills install all`
  61. (or `--all`) install every skill in the catalog.
  62. You can use `uvx` or `pixi exec` to install `skore` and run the installer in
  63. an isolated environment:
  64. ```bash
  65. uvx --from skore skore skills install
  66. ```
  67. or
  68. ```bash
  69. pixi exec --spec skore skore skills install
  70. ```
  71. If you prefer `npx`, then you can use:
  72. ```bash
  73. npx skills add probabl-ai/skore-skills
  74. ```
  75. ### Alternative — Claude Code plugin marketplace
  76. If you only use Claude Code and prefer the native plugin flow, this repo is
  77. also a [Claude Code plugin marketplace](https://docs.claude.com/en/docs/claude-code/plugin-marketplaces):
  78. ```bash
  79. /plugin marketplace add probabl-ai/skore-skills
  80. ```
  81. ```bash
  82. /plugin install probabl-skills@probabl-skills
  83. ```
  84. `/plugin update` pulls new releases.
  85. ## Skills in detail
  86. ### Meta and setup actions
  87. | Skill | Description |
  88. | --- | --- |
  89. | [triage-ml-task](skills/triage-ml-task/SKILL.md) | Session owner: list installed entry skills and ask which to run. |
  90. | [review-ml-choices](skills/review-ml-choices/SKILL.md) | Show stored project choices and re-enter the skill that can change one. |
  91. | [review-ml-experiment](skills/review-ml-experiment/SKILL.md) | Read the stored report, then write one idea file and one Ideas row per candidate. |
  92. | [setup-ml-project](skills/setup-ml-project/SKILL.md) | Ask only the pending setup pieces. When Git is selected, commit the current files before setup writes only if that check finds files to preserve. |
  93. | [setup-workspace](skills/setup-workspace/SKILL.md) | Detect or scaffold the standard ML workspace layout. |
  94. | [setup-python-env](skills/setup-python-env/SKILL.md) | Detect the env manager and bootstrap the runtime, agent-tools, and composed development environments. |
  95. | [setup-git](skills/setup-git/SKILL.md) | Initialize safe version control, including an optional commit of the tree before setup changes it. |
  96. | [persist-ml-git](skills/persist-ml-git/SKILL.md) | Commit the current loop stage when the end-turn hook says invoke. |
  97. | [model-ml-pipeline](skills/model-ml-pipeline/SKILL.md) | Require locked problem framing, write the first baseline design note for approval, then coordinate later choices, build, smoke testing, evaluation, and audit. |
  98. | [export-ml-project](skills/export-ml-project/SKILL.md) | Coordinate executed notebooks and an offline MkDocs documentation site. |
  99. | [sync-ml-reports](skills/sync-ml-reports/SKILL.md) | Copy skore reports between local, Hub, and MLflow, and optionally switch the upload destination. |
  100. ### ML pipeline lifecycle
  101. | Skill | Description |
  102. | --- | --- |
  103. | [explore-ml-data](skills/explore-ml-data/SKILL.md) | Explore the dataset before designing any model. |
  104. | [frame-ml-problem](skills/frame-ml-problem/SKILL.md) | Record the problem, deployment setting, metric, baseline, and fold count before any model code. |
  105. | [research-ml-practice](skills/research-ml-practice/SKILL.md) | Literature research for an ML methodology concern. |
  106. | [build-ml-pipeline](skills/build-ml-pipeline/SKILL.md) | Declare a skrub DataOps graph from the data source to the predictor after framing is locked. |
  107. | [evaluate-ml-pipeline](skills/evaluate-ml-pipeline/SKILL.md) | Evaluate one sklearn-compatible learner with the locked validation scheme and persist structured skore reports. |
  108. | [smoke-test-ml-pipeline](skills/smoke-test-ml-pipeline/SKILL.md) | Structural pytest: prediction count must match the predict-grid row count. |
  109. | [audit-ml-pipeline](skills/audit-ml-pipeline/SKILL.md) | Audit a persisted skore report read-only and produce a reusable digest. |
  110. ### Ideas and backlog
  111. | Skill | Description |
  112. | --- | --- |
  113. | [manage-ml-backlog](skills/manage-ml-backlog/SKILL.md) | Record experiment outcomes and triage idea files. A promoted idea moves from the Ideas table into a Backlog row. |
  114. | [shape-user-idea](skills/shape-user-idea/SKILL.md) | Shape a user idea or a named artifact into one idea file and one Ideas row after they confirm. |
  115. | [search-ml-literature](skills/search-ml-literature/SKILL.md) | Search scientific and technical sources and write one idea file and one Ideas row for the direction the user confirms. |
  116. ### Workspace and tooling
  117. | Skill | Description |
  118. | --- | --- |
  119. | [add-python-package](skills/add-python-package/SKILL.md) | Add a dependency through the project env manager, or ask the user when the environment is user-managed. |
  120. | [choose-python-library](skills/choose-python-library/SKILL.md) | Resolve a library choice and add the selected dependency. |
  121. | [plot-ml-figure](skills/plot-ml-figure/SKILL.md) | Pick pandas, seaborn, plotly, or matplotlib before writing figure code. |
  122. | [export-ml-notebook](skills/export-ml-notebook/SKILL.md) | Write a jupytext percent file out as a notebook with empty outputs. |
  123. | [export-ml-site](skills/export-ml-site/SKILL.md) | Package workspace markdown and existing notebook HTML into an offline MkDocs site. |
  124. Canonical package policy lives in the CLI; print it with `skore skills tools env stack` (the same command as `python -m skore_skills env stack`). `choose-python-library` resolves competing libraries.
  125. When executed notebooks or the documentation site are enabled in a
  126. workspace, workflow stages publish them at completed checkpoints before
  127. the next user decision. Python percent files execute once: that run can
  128. write the notebook, optional notebook HTML, and audit digest together.
  129. Site checkpoints rebuild only when Markdown, notebook HTML, figures,
  130. or result files changed. Unset or disabled notebook/site policies do
  131. not enable these artifacts automatically.
  132. ## Website
  133. The getting-started page lives in `site/` and is published at
  134. <https://probabl-ai.github.io/skore-skills/>.
  135. Build it with the `site` Pixi environment, which provides Node.js:
  136. ```bash
  137. pixi run -e site site-build
  138. ```
  139. Pull requests that change `site/`, `.catalog.json`, or the Pixi manifest build
  140. the site in CI. GitHub Pages builds it from `main` when those same paths change.
  141. A repo admin turns this on once: Settings → Pages → Source **GitHub Actions**.
  142. If Pages is enabled after that workflow is already on `main`, run the **Pages**
  143. workflow with **Run workflow**. Catalog edits still go through `pixi run check`.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.