Skip to main content
Agent skill

Python Project Structure

Python project organization, module architecture, and public API design. Use when setting up new projects, organizing modules, defining public interfaces with all, or planning directory layouts.

Get started

Get the complete skill folder, including its scripts and reference files.

When to use this skill

Use Python Project Structure when an AI agent needs a reusable SKILL.md workflow for this job: Python project organization, module architecture, and public API design. Use when setting up new projects, organizing modules, defining public interfaces with all, or planning directory layouts.

When not to use it

Skip Python Project Structure when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.

How to install

Where will you use it?

Install scope
Claude Code folder setup
mkdir -p ~/.claude/skills/python-project-structure
# Save SKILL.md to ~/.claude/skills/python-project-structure/SKILL.md
  1. Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/python-project-structure/SKILL.md. For a shared project, commit the folder at .claude/skills/python-project-structure/ instead.
  2. Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
  3. For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.

Full install guide for Claude, Cursor, and Codex →

Try a starter prompt

After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.

Starter prompt
Use the Python Project Structure skill for this task: Starting a new Python project from scratch. Confirm any required inputs or files before making changes.

What supports this recommendation?

90/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.

Documentation
100/100
Practical detail
100/100
Source evidence
63/100
Source signals
80/100
Model compatibility evidence
  • claudeuntested

    No model-specific signal or recorded compatibility test was found.

  • chatgptuntested

    No model-specific signal or recorded compatibility test was found.

  • geminiuntested

    No model-specific signal or recorded compatibility test was found.

  • copilotuntested

    No model-specific signal or recorded compatibility test was found.

  • llamauntested

    No model-specific signal or recorded compatibility test was found.

  • perplexityuntested

    No model-specific signal or recorded compatibility test was found.

  • mistraluntested

    No model-specific signal or recorded compatibility test was found.

  • grokuntested

    No model-specific signal or recorded compatibility test was found.

Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.

What this skill does

# Python Project Structure & Module Architecture

Design well-organized Python projects with clear module boundaries, explicit public interfaces, and maintainable directory structures. Good organization makes code discoverable and changes predictable.

## When to Use This Skill

- Starting a new Python project from scratch - Reorganizing an existing codebase for clarity - Defining module public APIs with `__all__` - Deciding between flat and nested directory structures - Determining test file placement strategies - Creating reusable library packages

## Core Concepts

### 1. Module Cohesion

Group related code that changes together. A module should have a single, clear purpose.

### 2. Explicit Interfaces

Define what's public with `__all__`. Everything not listed is an internal implementation detail.

### 3. Flat Hierarchies

Prefer shallow directory structures. Add depth only for genuine sub-domains.

### 4. Consistent Conventions

Apply naming and organization patterns uniformly across the project.

## Quick Start

``` myproject/ ├── src/ │ └── myproject/ │ ├── __init__.py │ ├── services/ │ ├── models/ │ └── api/ ├── tests/ ├── pyproject.toml └── README.md ```

## Fundamental Patterns

### Pattern 1: One Concept Per File

Each file should focus on a single concept or closely related set of functions. Consider splitting when a file:

- Handles multiple unrelated responsibilities - Grows beyond 300-500 lines (varies by complexity) - Contains classes that change for different reasons

```python # Good: Focused files # user_service.py - User business logic # user_repository.py - User data access # user_models.py - User data structures

# Avoid: Kitchen sink files # user.py - Contains service, repository, models, utilities... ```

### Pattern 2: Explicit Public APIs with `__all__`

Define the public interface for every module. Unlisted members are internal implementation details.

```python # mypackage/services/__init__.py from .user_service import UserService from .order_service import OrderService from .exceptions import ServiceError, ValidationError

__all__ = [ "UserService", "OrderService", "ServiceError", "ValidationError", ]

# Internal helpers remain private by omission # from .internal_helpers import _validate_input # Not exported ```

### Pattern 3: Flat Directory Structure

Prefer minimal nesting. Deep hierarchies make imports verbose and navigation difficult.

``` # Preferred: Flat structure project/ ├── api/ │ ├── routes.py │ └── middleware.py ├── services/ │ ├── user_service.py │ └── order_service.py ├── models/ │ ├── user.py │ └── order.py └── utils/ └── validation.py

# Avoid: Deep nesting project/core/internal/services/impl/user/ ```

Add sub-packages only when there's a genuine sub-domain requiring isolation.

### Pattern 4: Test File Organization

Choose one approach and apply it consistently throughout the project.

**Option A: Colocated Tests**

``` src/ ├── user_service.py ├── test_user_service.py ├── order_service.py └── test_order_service.py ```

Benefits: Tests live next to the code they verify. Easy to see coverage gaps.

**Option B: Parallel Test Directory**

``` src/ ├── services/ │ ├── user_service.py │ └── order_service.py tests/ ├── services/ │ ├── test_user_service.py │ └── test_order_service.py ```

Benefits: Clean separation between production and test code. Standard for larger projects.

## Advanced Patterns

### Pattern 5: Package Initialization

Use `__init__.py` to provide a clean public interface for package consumers.

```python # mypackage/__init__.py """MyPackage - A library for doing useful things."""

from .core import MainClass, HelperClass from .exceptions import PackageError, ConfigError from .config import Settings

__all__ = [ "MainClass", "HelperClass", "PackageError", "ConfigError", "Settings", ]

__version__ = "1.0.0" ```

Consumers can then import directly from the package:

```python from mypackage import MainClass, Settings ```

### Pattern 6: Layered Architecture

Organize code by architectural layer for clear separation of concerns.

``` myapp/ ├── api/ # HTTP handlers, request/response │ ├── routes/ │ └── middleware/ ├── services/ # Business logic ├── repositories/ # Data access ├── models/ # Domain entities ├── schemas/ # API schemas (Pydantic) └── config/ # Configuration ```

Each layer should only depend on layers below it, never above.

### Pattern 7: Domain-Driven Structure

For complex applications, organize by business domain rather than technical layer.

``` ecommerce/ ├── users/ │ ├── models.py │ ├── services.py │ ├── repository.py │ └── api.py ├── orders/ │ ├── models.py │ ├── services.py │ ├── repository.py │ └── api.py └── shared/ ├── database.py └── exceptions.py ```

## File and Module Naming

### Conventions

- Use `snake_case` for all file and module names: `user_repository.py` - Avoid abbreviations that obscure meaning: `user_repository.py` not `usr_repo.py` - Match class names to file names: `UserService` in `user_service.py`

### Import Style

Use absolute imports for clarity and reliability:

```python # Preferred: Absolute imports from myproject.services import UserService from myproject.models import User

# Avoid: Relative imports from ..services import UserService from . import models ```

Relative imports can break when modules are moved or reorganized.

## Best Practices Summary

1. **Keep files focused** - One concept per file, consider splitting at 300-500 lines (varies by complexity) 2. **Define `__all__` explicitly** - Make public interfaces clear 3. **Prefer flat structures** - Add depth only for genuine sub-domains 4. **Use absolute imports** - More reliable and clearer 5. **Be consistent** - Apply patterns uniformly across the project 6. **Match names to content** - File names should describe their purpose 7. **Separate concerns** - Keep layers distinct and dependencies flowing one direction 8. **Document your structure** - Include a README explaining the organization

Intended uses

  • Starting a new Python project from scratch
  • Reorganizing an existing codebase for clarity
  • Defining module public APIs with all
  • Deciding between flat and nested directory structures
  • Determining test file placement strategies
  • Creating reusable library packages

Implementation notes

  • Keep files focused - One concept per file, consider splitting at 300-500 lines (varies by complexity)
  • Define all explicitly - Make public interfaces clear
  • Prefer flat structures - Add depth only for genuine sub-domains
  • Use absolute imports - More reliable and clearer
  • Be consistent - Apply patterns uniformly across the project
  • Match names to content - File names should describe their purpose
  • Separate concerns - Keep layers distinct and dependencies flowing one direction
  • Document your structure - Include a README explaining the organization

Try it, then tell us how it went

A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.