Agent skills for Claude Desktop

Enforces strict Python code quality standards using pytest, mypy, ruff, black, and isort.

Python code style, linting, formatting, naming conventions, and documentation standards.

Python configuration management via environment variables and typed settings.

When to activate: ETL, data pipeline, Prefect, Airflow, DLT, Polars, batch processing, data quality

Python design patterns including KISS, Separation of Concerns, Single Responsibility, and composition over inheritance.

Python error handling patterns including input validation, exception hierarchies, and partial failure handling.

Enforces Python best practices, type safety, and modern development standards for robust backend applications.

When to activate: Strawberry, Graphene, GraphQL, subscriptions, dataloaders, N+1, schema-first design

When to activate: Python profiling, cProfile, memory profiling, optimization, numba, Cython, bottlenecks

You are a Python project architecture expert specializing in scaffolding production-ready Python applications.

Python project organization, module architecture, and public API design.

Python resilience patterns including automatic retries, exponential backoff, timeouts, and fault-tolerant decorators.

Python resource management with context managers, cleanup patterns, and streaming.

Facilitates the creation of high-performance Python bindings for Rust cores using PyO3 and idiomatic wrapper patterns.

Architects modular, maintainable standalone Python scripts and CLI utilities using a structured ports-and-adapters pattern optimized for…

Audits Python libraries for security vulnerabilities using Bandit, pip-audit, Semgrep, and detect-secrets.

Python type safety with type hints, generics, protocols, and strict type checking.

When to activate: Python type hints, TypeVar, Generic, Protocol, TypedDict, Literal, typing extensions, mypy

When to activate: WebSockets, real-time, FastAPI WebSocket, websockets library, broadcasting, connection management

When to activate: PyTorch, nn.Module, DataLoader, training loop, mixed precision, DDP, distributed training, torch.compile, model export…

Use when writing, editing, reviewing, or debugging q/kdb+ code (.q files), querying kdb+ tables, translating Python to q, running q from…