원클릭으로 Manus에서 모든 스킬 실행

python-code-style

스타2

포크1

업데이트2026년 5월 2일 12:01

Owns Python code style for this stack: ruff for lint + format, numpydoc for docstrings. Two responsibilities — (1) place the project's `ruff.toml` from the bundled template once the stack and workspace are in place, and (2) run ruff against any Python files Claude has just generated or edited. Stops at "the touched files pass `ruff check`." TRIGGER when (any of these): (1) a Python file was just created or edited via Write / Edit / MultiEdit — invoke this skill before declaring the task done so ruff is run on the touched files; (2) a fresh ML workspace was just scaffolded by `organize-ml-workspace` and the project has no `ruff.toml` at its root yet — drop the bundled template; (3) the user asks about lint, format, docstring style, or reaches for `black` / `isort` / `flake8` / `pydocstyle` (redirect to ruff — the stack's canonical linter, owned by `data-science-python-stack` Tier 1). SKIP when: the project is non-Python; the only edits in this turn are to Markdown / TOML / JSON / Y

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

Manus에서 실행

출처

glemaitre

glemaitre/experiments

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Manus에서 실행

관련 직업SOC

SOC 직업 분류 기준

소프트웨어 개발자컴퓨터 및 수학직·SOC 15-1252

파일 탐색기

2 개 파일

SKILL.md

readonly

이 저장소의 다른 Skills

같은 저장소

build-ml-pipeline

glemaitre/experiments

Opinionated, Pythonic way to **declare** the pipeline that goes from a data source to a predictor: data loading, preprocessing, feature engineering, estimator selection, and their composition. The pipeline is built as a **skrub DataOps graph**; every step is either a pure-Python function (stateless) attached via `.skb.apply_func`, or a scikit-learn-compatible estimator (stateful) attached via `.skb.apply`. Stops at the declared object. Out of scope: `fit` invocation, train/test split, hyperparameter tuning, persistence, evaluation. Deep-learning declarations are covered via internal `references/*.md`; skrub and scikit-learn mechanics live in sibling skills. TRIGGER when: writing or editing code that declares any link in the chain *data source → predictor* — data readers/loaders feeding a model (`read_csv`, `read_parquet`, `Dataset` classes), preprocessing or feature-engineering steps (transformers, encoders, imputers, scalers, text/image featurizers), **pure-Python data-processing functions destined for the p

2026-05-032

organize-ml-workspace

glemaitre/experiments

Decide where files live in an ML experimentation project: where the reusable code goes, where each experiment goes, where reports are persisted. Owns the layout, the file-creation rules (one file per experiment, ask before editing an existing one), and the jupytext-style `# %%` script convention. Stops at "the file exists in the right place with the right skeleton". Never imposes a `data/` layout — the user owns that. TRIGGER when: starting a new ML project / scaffolding a workspace; about to create the first experiment file in a project; about to create `src/<pkg>/data.py`, `features.py`, `pipeline.py`, or `evaluate.py` for the first time; about to write a Jupyter notebook (`.ipynb`) for experimentation — redirect to a `# %%` script under `experiments/`; user asks where something should live, how to organize the project, or how to set up the workspace; about to add a *new experiment iteration* (must decide: new file vs. edit existing — ask the user). SKIP when: the file is clearly part of the package's exist

2026-05-032

python-env-manager

glemaitre/experiments

Single source of truth for "which Python environment manager does this project use, and how do I install a package with it?". Owns the detection table (pixi / uv / poetry / hatch / conda+mamba / pip+venv), the install / remove / upgrade commands per manager, and the bootstrap path when no manager is in place (default recommendation: pixi). Stops at "the install command was issued with the right manager and the package is importable". TRIGGER when (any of these): (1) **about to install / add / pin / upgrade / remove a Python package** — `pip install`, `pixi add`, `uv add`, `poetry add`, `conda install`, etc. — under any framing; (2) `data-science-python-stack` § "Missing dependency" surfaced a missing import and an install is the next step; (3) a workflow skill's Stop condition fired on a missing dependency (`build-ml-pipeline`, `evaluate-ml-pipeline`, `organize-ml-workspace`); (4) starting a new Python project and no manager is in place yet (bootstrap with pixi unless the user picks ot

2026-05-032

iterate-from-diagnostic

glemaitre/experiments

Source the next ML experiment proposal by inspecting the skore report from the previous run — residuals, calibration, per-slice metrics, threshold behavior, statistical checks surfaced by skore. Hand the proposal back to `iterate-ml-experiment`, which writes it into `plan/NN_short_name.md` and seeks the user's approval. Stops at "a proposal (question, motivation, method outline) has been returned"; does not write any plan file itself, and does not author acceptance criteria — the user judges the result. TRIGGER when: `iterate-ml-experiment` is picking a sourcing strategy, the previous experiment has a skore report on disk, and the user says "the report shows X", "calibration looks bad", "why is slice Y so off?", "residuals look weird", or "what does the report tell us?"; the user is open-ended after a recent run — try this strategy first when a fresh report exists. SKIP when: the previous experiment hasn't run yet (no report to read); the user has a concrete modelling idea (use `iterate-from-user`); the issue

2026-05-022

iterate-from-literature

glemaitre/experiments

Source the next ML experiment proposal by searching the scientific / engineering literature (papers, blog posts, library docs) for techniques applicable to the current problem. Hand the proposal back to `iterate-ml-experiment`, which writes it into `plan/NN_short_name.md` and seeks the user's approval. Stops at "a proposal (question, motivation, method outline, source citations) has been returned"; does not write any plan file itself, and does not author acceptance criteria — the user judges the result. TRIGGER when: `iterate-ml-experiment` is picking a sourcing strategy and the user asks "any papers on this?", "what does the literature say?", "how do people usually handle X?", or "search for techniques for Y"; the user has run out of obvious next steps and is open-ended; a sibling strategy (methodology / diagnostic) has surfaced a problem with no obvious in-house fix and external prior art would help. SKIP when: the user has a concrete idea already (use `iterate-from-user`); the gap is clearly methodological

2026-05-022

iterate-from-methodology

glemaitre/experiments

Source the next ML experiment proposal by auditing the *methodology* of the previous experiment(s) — split strategy, leakage risk, target encoding, sample size, metric choice, baseline comparability, randomness control. Hand the proposal back to `iterate-ml-experiment`, which writes it into `plan/NN_short_name.md` and seeks the user's approval. Stops at "a proposal (question, motivation, method outline) has been returned"; does not write any plan file itself, and does not author acceptance criteria — the user judges the result. TRIGGER when: `iterate-ml-experiment` is picking a sourcing strategy and the user says "did we get the split right?", "is this leaking?", "small sample size?", "is the baseline fair?", "is this metric the right one?"; the previous experiment's result looks suspicious (too good, too noisy, too flat) and the user wants to check the setup before iterating further; a literature / diagnostic strategy has surfaced something that turns out to be a methodology issue, not a modelling issue. SKI

2026-05-022

name

python-code-style

description

Python Code Style

Single owner of "what does well-styled Python look like in this stack": ruff (lint + format) and numpydoc docstrings. This skill is explicitly manual — Claude runs ruff on the files it has just touched, no hook involved.

Stop conditions — read before anything else

Do not configure a PostToolUse / PreToolUse hook for ruff. This skill is intentionally manual. A hook tightens the loop in ways that bite (every micro-edit triggers a fix cycle, partial files fail D-rule checks mid-write, retries can stall the turn). If the user explicitly asks for an automated hook later, redirect to update-config — but the default is "Claude runs ruff itself."
Do not substitute ruff with black / isort / flake8 / pydocstyle / pylint. Ruff is the canonical linter in this stack (data-science-python-stack Tier 1). If import ruff / pixi run ruff --version fails, route through python-env-manager to install — don't silently fall back.
One fix attempt per file, then surface. If ruff check reports issues after Claude's first fix, address them once. If the same issue persists after the second pass, stop editing that file and surface the remaining diagnostics + diff to the user. This is the anti-infinite-loop guardrail — do not enter a third cycle on the same warning.
Don't lint files outside the user's code. The hook scope is src/<pkg>/, experiments/, top-level *.py scripts, and any package directory the user owns. Skip vendored paths, generated files, and anything under .pixi/, .venv/, node_modules/, etc.

Pre-flight — emit this checklist as visible text before running ruff

Pre-flight (python-code-style):
- [ ] ruff importable in the project's env (`pixi run ruff --version`
      succeeds, per `data-science-python-stack` Tier 1)
- [ ] `ruff.toml` present at project root — if absent and the stack
      + workspace are already set up, drop `templates/ruff.toml`
      from this skill into the root before running ruff
- [ ] File list ready: <abs paths of .py files touched this turn>
- [ ] Decision recorded: this is the first ruff pass on these files
      (proceed) | second pass (proceed but stop on persistent
      issues) | third pass on same warning (STOP, surface to user)

Scope

In scope: running ruff format + ruff check --fix + ruff check on Python files Claude has just generated or edited; authoring numpydoc docstrings on public functions and classes; dropping the ruff.toml template into a fresh project.
Out of scope: type hints (mypy / pyright are not in the stack); naming conventions ruff doesn't enforce; setting up PostToolUse / PreToolUse hooks; linting non-Python files.

What to run, in what order

For every Python file touched this turn (call them <files>), run inside the project's environment manager — pixi run for pixi projects, equivalent for uv / poetry / conda (per python-env-manager):

pixi run ruff format <files>
pixi run ruff check --fix <files>
pixi run ruff check <files>

Three steps, in order:

ruff format — applies the formatter (line length, quoting, trailing commas, blank lines around defs). Idempotent.
ruff check --fix — auto-fixes everything ruff knows how to fix in place: import sorting (I), legacy syntax (UP), detectable bug patterns (B).
ruff check (no --fix) — final pass. Anything reported here needs Claude's attention: missing docstrings (D), undefined names (F), code structure issues. Address them, then re-run the trio. Apply the one-fix-per-file rule from Stop conditions.

If a file under experiments/ has a D100 ("missing module docstring") or D103 ("missing function docstring") warning, that's expected for # %% cells; the bundled ruff.toml ignores those for experiments/**. If you're seeing them, the ruff.toml isn't loaded — check that it lives at the project root.

Numpydoc — the docstring convention

Public functions and classes carry numpydoc-format docstrings; ruff's D rules with pydocstyle.convention = "numpy" enforce the shape. Skeleton:

def predict_price(X, model, *, n_jobs=1):
    """Predict option prices from a feature matrix.

    Parameters
    ----------
    X : pandas.DataFrame
        Feature matrix with one row per option.
    model : sklearn.base.BaseEstimator
        Fitted estimator with a ``predict`` method.
    n_jobs : int, default=1
        Number of parallel jobs.

    Returns
    -------
    numpy.ndarray of shape (n_samples,)
        Predicted prices, one per row of ``X``.
    """

Conventions worth surfacing because they're non-obvious:

One-line summary on the first line, in the imperative mood ("Predict ..." not "Predicts ..."). No trailing period in the summary line if D400 is enabled — but in numpy convention it is, so write the period.
Blank line between summary and the rest.
Parameter shapes go in the type slot, not the description, e.g. X : ndarray of shape (n_samples, n_features).
Returns section lists the return value; if there are multiple returns, list each on its own row. Don't omit the type.
Private helpers (_leading_underscore) don't need a docstring under the default D rules — ruff allows that.
Modules (top of file) should start with a one-line summary. Skipped under experiments/** per the bundled ruff.toml.

Initial setup — dropping the `ruff.toml` template

When this skill is invoked on a fresh project that has no ruff.toml at its root and the stack + workspace have been scaffolded by their respective skills:

Read templates/ruff.toml from this skill (path: .claude/skills/python-code-style/templates/ruff.toml).
Write it verbatim to <project-root>/ruff.toml.
Verify ruff picks it up: pixi run ruff check --show-settings . should report the numpy convention and the select list from the template.

Do not fold ruff config into pyproject.toml automatically — the project may not have one, or the user may prefer a separate file. The standalone ruff.toml is unambiguous.

When ruff finds something Claude didn't write

A common case: Claude edits one function in a file that already had unrelated D-rule violations. Ruff will report those too.

In scope of this turn: the lines Claude touched. Fix those.
Out of scope: pre-existing warnings in untouched code. Mention them in the response so the user can choose to address them, but don't drag every warning into the current task.

This keeps PR scope tight and avoids "while I was here" expansion that the user didn't ask for.

Companion skills

data-science-python-stack — owns the decision that ruff is Tier 1 mandatory; this skill assumes ruff is already installed. If pixi run ruff --version fails, return there for the install.
python-env-manager — turns "ruff is missing" into the right install command for the project's manager. Don't run pip install ruff in a pixi project.
organize-ml-workspace — sets up the directory layout that the bundled ruff.toml's per-file ignores (experiments/**) reference. Drop the template after this skill has run, so the paths it ignores actually exist.
update-config — only relevant if the user explicitly asks for an automated lint hook later. Default is no hook.

Conventions

Manual, not automatic. Claude calls ruff itself, file by file. No hook.
One-fix-per-file rule. Hard cap on retries — second pass max, then surface.
Project-root config only. ruff.toml lives at the project root and is the single source of truth. Don't add per-directory overrides unless the user asks.
Don't widen scope on touched files. Fix what Claude wrote; surface (don't auto-fix) pre-existing issues elsewhere.

python-code-style

이 저장소의 다른 Skills

이 저장소의 다른 Skills

Python Code Style

Stop conditions — read before anything else

Pre-flight — emit this checklist as visible text before running ruff

Scope

What to run, in what order

Numpydoc — the docstring convention

Initial setup — dropping the ruff.toml template

When ruff finds something Claude didn't write

Companion skills

Conventions

Python Code Style

Stop conditions — read before anything else

Pre-flight — emit this checklist as visible text before running ruff

Scope

What to run, in what order

Numpydoc — the docstring convention

Initial setup — dropping the ruff.toml template

When ruff finds something Claude didn't write

Companion skills

Conventions

Initial setup — dropping the `ruff.toml` template

Initial setup — dropping the `ruff.toml` template