coding-guidance-python（coding guidance Python 搜索）

Python Coding Guidance

This skill adds portable Python implementation, refactoring, and review guidance.

Adjacent Skills

This skill provides portable Python engineering principles. Compose with:

• Workflow: thinking (planning), recursive-thinking (stress-testing), security (threat modeling)
• Domain overlays: backend-guidance (server-side code), backend-systems-guidance (stronger backend architecture, reliability, and trust-boundary work), ui-guidance (graphical UI/web frontend), project-core-dev (repo-specific build/test commands)

Open the bundled references only when the task actually touches package distribution, entrypoint layout, or a real I/O/service boundary.

When Not to Lean on This Skill

• non-Python work
• notebook-first exploratory work where reproducibility and module design are not the main concern
• one-off throwaway scripts where the repo explicitly does not want production Python standards

Implementation Workflow

1. Read the touched modules, entrypoints, tests, package metadata, and nearby docs before editing.
2. Infer intended behavior from existing code, imports, and tests when the request is only partially specified. Ask only when multiple plausible designs would change semantics.
3. Choose the narrowest change that keeps contracts, side effects, error handling, and public API shape explicit.
4. Implement with simple functions, clear module boundaries, explicit types where they improve the contract, and production-safe behavior at I/O boundaries.
5. Add or update tests close to the changed behavior; prefer pytest-native fixtures and parameterization over ad hoc setup.
6. Run the narrowest relevant formatter, linter, type checker, packaging check, and test targets the repo supports.

Refactoring Workflow

Use this instead of the default implementation workflow when the task is primarily cleanup or restructuring:

1. Capture current behavior, side effects, hidden globals, import shape, and mutation hotspots.
2. Break the refactor into small slices that preserve behavior.
3. Remove long functions, muddled responsibilities, implicit coupling, and anti-patterns one step at a time.
4. Keep tests passing after each slice; add characterization coverage first when behavior is unclear.
5. Stop when the code is simpler, more explicit, easier to test, and easier to operate.

Review Workflow

When reviewing (not implementing), skip the implementation workflow and use this instead:

1. Read the change in full before commenting.
2. Identify findings, ordered by severity: Critical > Important > Suggestion.
3. Prioritize bugs and regressions, exception or error-path holes, mutable shared-state mistakes, typing and interface mismatches, packaging or import breakage, performance mistakes with real impact, security risks, and missing tests. Add service-boundary observability, retry, and timeout findings when the change actually crosses a real external boundary.
4. State findings with concrete evidence and the likely consequence.

Python Rules

First tier - causes bugs

• Keep module side effects minimal; avoid import-time network calls, filesystem mutation, or heavy initialization unless the module is explicitly an entrypoint
• Prefer explicit parameters and return values over hidden globals, ambient context, or mutation of module-level state
• Do not use mutable default arguments
• Treat None handling, optional fields, and missing keys as contract design, not cleanup for callers to guess
• Preserve exception context; do not catch broadly and discard the original failure without adding useful domain context
• Use context managers for files, locks, subprocess pipes, and other resources with lifetime rules
• Be explicit about text vs. bytes boundaries, timezone-aware vs. naive datetimes, and sync vs. async call paths
• Do not block the event loop with synchronous I/O in async code
• Do not mix sync and async APIs inside one path without a clear boundary

Second tier - prevents mistakes

• Prefer small functions and plain data flow before introducing classes
• Use classes when they model stateful domain objects or a stable behavior boundary, not just to group helpers
• Prefer pathlib.Path over stringly typed path manipulation
• Prefer dataclass, TypedDict, Protocol, Enum, and other standard library types when they make contracts clearer
• Use comprehensions and built-ins when they make intent clearer; avoid dense one-liners that hide control flow
• Avoid boolean flag parameters when separate functions or a small config type would make behavior clearer
• Keep warnings, lint findings, and type-checker regressions at zero in repo-owned code

Typing and interfaces

• Add type hints to public functions, methods, and non-trivial internal seams where they clarify real contracts
• Prefer concrete types at boundaries and protocols for substitution seams rather than broad Any
• Use Any only when the repo truly needs a dynamic escape hatch and the cost is explicit
• Prefer modern union syntax and type narrowing over comments or sentinel-heavy calling conventions
• Public APIs should make invalid states hard to represent and expensive work visible to callers
• If a function takes more than 2-3 meaningful parameters, prefer a named type, config object, or split responsibility
• Make mutation visible in names and interfaces; hidden in-place updates are a design smell
• Use strict type checking where the repo supports it; if the codebase is not yet strict, tighten changed modules instead of broadening exemptions

Tests and verification

• Prefer pytest-style tests with clear arrange-act-assert flow
• Use fixtures for reusable setup and teardown, not for hiding the meaning of a test
• Use parameterized tests when the same contract should hold across multiple inputs
• Mock external boundaries, not the internal behavior you are trying to prove
• Add integration coverage when the change crosses package, transport, persistence, or subprocess boundaries
• Coverage is a signal, not a goal; prioritize meaningful path coverage over percentage chasing

Data and state discipline

• Keep validation, serialization, persistence, and business rules separated enough that each can be tested directly
• Be suspicious of dict-shaped data flowing through many layers without a named contract
• Prefer immutable or append-only data flow where shared mutation would make behavior harder to reason about
• Cache only when measurement or repeated cost justifies it; make cache scope and invalidation rules explicit
• Treat environment-variable reads, current working directory assumptions, and process-global configuration as boundary concerns
• Validate external input at the boundary instead of letting malformed data fail deep in the call graph

Modules, structure, and packaging

• Prefer narrow modules with clear import direction over utility grab bags
• Avoid circular imports by moving shared contracts to a lower layer instead of using late imports as a default escape hatch
• Keep CLI, transport, persistence, and domain logic separated enough to test each piece in isolation
• Prefer shallow directory structures unless there is a real sub-domain split
• Expose public package APIs deliberately; do not leak internal helpers, framework types, or persistence models as public contracts by accident
• Packaging-specific guidance that would otherwise bloat this file lives in references/python-packaging-and-layout.md

Architecture and anti-patterns

• Prefer composition over inheritance unless inheritance matches the real domain model
• Wait for repeated pressure before extracting abstractions; avoid clever registries, factories, or framework indirection that do not buy real clarity
• Keep I/O and business logic separated enough that core logic can be tested without transport or database setup
• Do not duplicate retry, timeout, or fallback behavior at multiple layers
• Do not hard-code secrets, production hosts, or environment-specific settings
• Do not swallow exceptions, ignore partial failures silently, or hide destructive side effects behind benign names

Concurrency, observability, and resilience

• Keep sync and async entrypoints separate unless the repo already has a clear bridging pattern
• In async code, await real work promptly; do not create background tasks casually without ownership, cancellation, and error propagation rules
• Protect invariants, not individual fields, when threads or async tasks share mutable state
• Service-boundary concerns become first-class only when the task actually crosses a real external boundary such as HTTP, queues, subprocesses, or persistent workers
• Service-boundary observability and resilience guidance lives in references/python-service-boundaries.md

Decision Heuristics

Use these when the right choice is not obvious:

• Scope check: if a change touches more than 3 modules or public entrypoints, stop and plan before continuing; the change is bigger than it looks.
• State visibility: if mutation or side effects are hard to see from the function signature, redesign the interface or add a one-line contract comment.
• Typing pressure: if Any, untyped dicts, or loose tuples start spreading, introduce a clearer type boundary before adding more code.
• Async boundary: if a change crosses sync and async code, name the boundary explicitly and keep adaptation local.
• Packaging pressure: if the change affects imports, entrypoints, package layout, or published metadata, treat that as a compatibility boundary rather than a refactor detail.
• I/O boundary pressure: if an external boundary can fail, time out, retry, or degrade, make the ownership and diagnostics explicit; local helper code usually does not need the same instrumentation.
• Repo conventions: if the repo has established formatter, linter, typing, or framework conventions, follow them unless they create a correctness or maintainability problem.
• Narrowness vs. quality: implement the narrowest change that solves the problem. When narrowness conflicts with correctness or clarity, prefer correctness. When it conflicts with style alone, prefer narrowness unless the task is explicitly a cleanup.
• Refactor boundary: outside explicit refactor work, fix at most one small adjacent issue while you are in the file.
• Abstraction threshold: three similar code blocks or repeated data-shaping pain is a pattern; before extracting, check whether a helper function, named type, or boundary cleanup is the simpler move.
• Performance rule: optimize only after measurement, except for obvious algorithmic, allocation, or I/O mistakes on hot paths.
• Framework pressure: if a framework convenience hides control flow, data ownership, or test seams, prefer the plainer construct.
• Language-fit check: if the problem is mostly shell orchestration, keep it in shell; if the problem needs rich data shaping, type-safe contracts, or non-trivial retry and observability logic, prefer Python over stretching shell too far.

Validation

A change is done when:

• the code passes the repo's formatter or format-check
• lint and static analysis report no new findings
• type checking reports no new regressions where the repo uses type checking
• existing tests pass
• new or changed behavior has test coverage, or the lack of coverage is called out with a concrete reason
• changed CLI, import, or service entrypoints have a narrow smoke path
• packaging metadata and import paths are verified when the change affects a distributable library or CLI
• review findings at Critical and Important severity are addressed

Examples

• Refactor this Python service module without breaking its typed public API
• Review this async Python worker for retry, timeout, and test gaps