Write maintainable Python
Use clear modules, functions, classes, dataclasses, type hints, exceptions, context managers, iterators, and standard-library tools.
Codex Skills for Python give Codex practical instructions for organizing packages, writing typed and readable Python, building scripts and services, handling asynchronous work, testing behavior, managing environments, and profiling performance. Download the skills for Python projects that need more than a quick script that only works on one machine.
task: complete python development task
inspect:
- requirements and context
- existing standards
- failure and edge cases
verify: outputs + checks + handoffStrong python development work depends on context, explicit decisions, and checks that match the real environment.
These skills give Codex a specialist workflow for python development while preserving local conventions, important constraints, and human review.
Use clear modules, functions, classes, dataclasses, type hints, exceptions, context managers, iterators, and standard-library tools.
Create command-line tools, services, data pipelines, file processors, integrations, workers, and scheduled scripts.
Handle asyncio, cancellation, timeouts, clients, files, databases, concurrency limits, cleanup, and retry behavior.
Use isolated environments, dependency metadata, unit and integration tests, linting, type checks, profiling, and distribution workflows.
The steps keep context, implementation, and verification visible so the result can be reviewed and repeated.
Read supported versions, pyproject settings, packages, entry points, dependencies, environments, tests, and local instructions.
Identify types, state, I/O, exceptions, resources, async boundaries, configuration, and callers.
Keep functions focused, types useful, dependencies justified, errors explicit, and resource lifetimes controlled.
Execute tests, type and lint checks, packaging validation, integration checks, and profiling where risk requires it.
The workflow adjusts to the project, audience, tools, and risk while preserving the same quality standard.
Create reliable scripts, commands, configuration, file handling, logs, errors, and distributable entry points.
Build APIs, validation, persistence, background work, authentication, testing, and operational behavior.
Organize transformations, notebooks, packages, reproducibility, validation, and performance.
Design public APIs, typing, documentation, examples, tests, versioning, and package publication.
Start with one defined outcome and provide the source material, constraints, and checks that matter.
These skills are designed for people who need dependable python development work with a visible process.
Build applications, services, libraries, automation, tests, and performance improvements.
Turn notebook logic into reusable, validated, documented, and reproducible Python modules.
Create dependable scripts and integrations with configuration, logs, retries, and safe failure behavior.
Understand language features, modules, environments, errors, testing, typing, and packaging.
The skills assist Python development, but production security, dependency trust, data handling, concurrency, and deployment require review and representative testing.
Install the complete skill folder and add the project-specific context before beginning.
Keep SKILL.md with the included python development guidance, checks, and supporting files.
Use project scope for local rules and versions, or personal scope for a reusable workflow.
Document the tools, audience, standards, source material, constraints, and approval rules for this python development work.
Provide the intended outcome, important edge cases, required checks, and expected handoff.
Practical answers about capabilities, limits, setup, and review.
The project metadata and environment define the supported interpreter versions and available features.
Yes. They follow frameworks already installed in the project and their version-specific conventions.
Yes. They can structure modules, pyproject metadata, entry points, tests, typing, documentation, and builds.
Yes. They can design tasks, cancellation, timeouts, resource cleanup, concurrency limits, and async tests.
They can update project metadata and lockfiles using the existing package workflow, then run compatibility checks.
Clear context. Purposeful work. Relevant checks. A result others can understand.