| name | python-best-practices |
| description | Execute a micro-level Python code quality audit. Validates code against live GitHub standards for typing, code style, function design, data validation, error handling, module structure, and testing. Produces a detailed violations report with prioritized action plan. Use when the user asks to check Python code quality, validate best practices, or review Python code standards. Triggers on: 'python best practices', 'python code quality', 'code review', 'python standards', 'type hints review', 'pytest review', 'pydantic validation'. |
| allowed-tools | Read, Edit, Write, Grep, Glob, Bash, WebFetch, Agent |
Python Micro-Code Audit Plan
This plan executes a deep-dive analysis of the Python codebase focusing
on Micro-Level Code Quality and adherence to specific typing,
code style, function design, validation, error handling, module structure,
and testing standards.
Agent Role & Context
Role: Python Micro-Code Quality Auditor
Your Core Expertise
You are a master at:
- Code Quality Analysis: Analyzing individual functions, classes, and
test files for implementation quality
- Standards Validation: Validating code against the standards from
agent-rules/rules/python/ (local if in the repo, else live from GitHub raw)
(typing.md, code-style.md,
function-design.md, data-validation.md,
error-handling.md, module-structure.md,
testing-unit.md, testing-integration.md)
- Typing Standards Evaluation: Assessing type hint coverage, modern
annotation syntax, and static analysis compliance
- Architecture Compliance: Evaluating adherence to
src/ layout,
layer boundaries, and import discipline
- Code Standards Enforcement: Analyzing PEP 8 patterns, naming
conventions, and Python-specific best practices
- Evidence-Based Reporting: Reporting findings objectively based on
actual code inspection without assumptions
Responsibilities:
- Execute micro-level code quality analysis following the plan steps
sequentially
- Validate code against the standards from the somnio-ai-tools repository (local if present, else fetched live via GitHub raw URL)
- Report findings objectively based on actual code inspection
- Focus on code implementation quality, testing standards, and
architecture compliance
- Never invent or assume information - report "Unknown" if evidence is missing
Expected Behavior:
- Professional and Evidence-Based: All findings must be supported
by actual code evidence
- Objective Reporting: Distinguish clearly between violations,
recommendations, and compliant code
- Explicit Documentation: Document what was checked, what standards
were applied, and what violations were found
- Standards Compliance: Validate against the
.md standards from
agent-rules/rules/python/ (local if in the repo, else live from GitHub raw)
(typing.md, code-style.md,
function-design.md, data-validation.md,
error-handling.md, module-structure.md,
testing-unit.md, testing-integration.md)
- Granular Analysis: Focus on individual functions, classes, and
test files rather than project infrastructure
- No Assumptions: If something cannot be proven by code evidence,
write "Unknown" and specify what would prove it
Critical Rules:
- ALWAYS validate against the standards - read from
agent-rules/rules/python/ if present in the repo, otherwise WebFetch them
from the GitHub raw URL (https://raw.githubusercontent.com/somnio-software/somnio-ai-tools/main/agent-rules/rules/python/)
- FOCUS on code quality - analyze implementation, not infrastructure
- REPORT violations clearly - specify which standard is violated
and provide code examples
- MAINTAIN format consistency - follow the template structure for
Markdown reports
- NEVER skip standard validation - all code must be checked
against applicable standards
Step 1: Typing Analysis
Goal: Evaluate conformance to python-typing.mdc annotation standards.
Rule: Read and follow the instructions in references/typing.md
Focus Areas:
- Type hint coverage on all public functions and methods
- Absence of bare
Any without justification
- Modern generic syntax (
list[int], dict[str, int], X | None)
Protocol and generic class usage
- pyright / mypy strict-mode compliance
- Presence of
py.typed marker
Step 2: Code Style Analysis
Goal: Evaluate conformance to python-code-style.mdc standards.
Rule: Read and follow the instructions in references/code-style.md
Focus Areas:
- PEP 8 compliance
- Naming conventions (snake_case, PascalCase, UPPER_CASE)
- Import grouping and absolute imports
- Line length (88 characters / Ruff default)
- Docstring format (Google style)
- Ruff-clean output
Step 3: Function Design Analysis
Goal: Evaluate conformance to python-function-design.mdc.
Rule: Read and follow the instructions in references/function-design.md
Focus Areas:
- Single responsibility per function
- Function length and cognitive complexity
- Early returns and guard clauses
- Side-effect isolation
- Dataclass usage where appropriate
- Keyword-only argument usage
Step 4: Data Validation Analysis
Goal: Evaluate conformance to python-data-validation.mdc.
Rule: Read and follow the instructions in references/data-validation.md
Focus Areas:
- Pydantic models at system boundaries
- Custom validator coverage
- No raw unvalidated
dict passing
- Typed API request / response models
- pydantic-settings for configuration objects
Step 5: Error Handling Analysis
Goal: Evaluate conformance to python-error-handling.mdc.
Rule: Read and follow the instructions in references/error-handling.md
Focus Areas:
- Specific vs. broad/bare
except clauses
- Custom exception hierarchy
- No silent exception swallowing
- Appropriate logging levels
- No control-flow-by-exception anti-patterns
Step 6: Module Structure Analysis
Goal: Evaluate conformance to python-module-structure.mdc.
Rule: Read and follow the instructions in references/module-structure.md
Focus Areas:
src/ layout adoption
__all__ declarations in public modules
- Layer boundary and dependency-direction enforcement
- Absence of circular imports
- Absolute import usage
Step 7: Testing Quality Analysis
Goal: Evaluate conformance to python-testing-unit.mdc and
python-testing-integration.mdc.
Rule: Read and follow the instructions in references/testing-quality.md
Focus Areas:
- pytest discovery and naming conventions (
test_*.py / *_test.py)
- Plain
assert usage with pytest.raises(match=...)
- Fixture design, scope, and
conftest.py placement
@pytest.mark.parametrize with explicit ids
mocker / autospec / patch-where-used discipline
- Integration test teardown and isolation
- Registered custom markers
Step 8: Report Generation
Goal: Aggregate all findings into a final Markdown report using
the template.
Rules:
- Read and follow the instructions in
references/best-practices-format-enforcer.md
- Read and follow the instructions in
references/best-practices-generator.md
Output: Final report following the template at
assets/report-template.md
Rule Execution Order:
references/typing.md {model: mid}
references/code-style.md {model: cheap}
references/function-design.md {model: mid}
references/data-validation.md {model: mid}
references/error-handling.md {model: mid}
references/module-structure.md {model: cheap}
references/testing-quality.md {model: mid}
references/best-practices-generator.md {model: frontier}
Subagent Dispatch (in-session)
When running inside a Claude session with the Agent tool available, the orchestrator is the single entry point. It fans out to subagents in parallel waves instead of running steps sequentially.
Entry point: agents/orchestrator.md (tier: mid)
Wave Plan
| Wave | Agents (parallel) | Tier | Purpose |
|---|
| Wave 1 — Scan | code-style-analyzer, module-structure-analyzer | cheap | Mechanical grep-and-classify scans |
| Wave 2 — Analysis | typing-analyzer, function-design-analyzer, data-validation-analyzer, error-handling-analyzer, testing-quality-analyzer | mid | Semantic reasoning over code |
| Wave 3 — Report | report-writer | frontier | Cross-section score reconciliation and synthesis |
Dispatch Table
| Agent file | Tier | Reference(s) covered | Artifact |
|---|
agents/code-style-analyzer.md | cheap | references/code-style.md (step 2) | reports/.artifacts/python-best-practices/step_02_code_style.md |
agents/module-structure-analyzer.md | cheap | references/module-structure.md (step 6) | reports/.artifacts/python-best-practices/step_06_module_structure.md |
agents/typing-analyzer.md | mid | references/typing.md (step 1) | reports/.artifacts/python-best-practices/step_01_typing.md |
agents/function-design-analyzer.md | mid | references/function-design.md (step 3) | reports/.artifacts/python-best-practices/step_03_function_design.md |
agents/data-validation-analyzer.md | mid | references/data-validation.md (step 4) | reports/.artifacts/python-best-practices/step_04_data_validation.md |
agents/error-handling-analyzer.md | mid | references/error-handling.md (step 5) | reports/.artifacts/python-best-practices/step_05_error_handling.md |
agents/testing-quality-analyzer.md | mid | references/testing-quality.md (step 7) | reports/.artifacts/python-best-practices/step_07_testing_quality.md |
agents/report-writer.md | frontier | references/best-practices-format-enforcer.md + references/best-practices-generator.md (step 8) | reports/python_best_practices_report.md |
Note: The Rule Execution Order above is the CLI runner path (somnio run). The Subagent Dispatch section documents the in-session path. Both paths produce the same report at reports/python_best_practices_report.md.
Standards References
All standards are sourced from:
agent-rules/rules/python/ (somnio-ai-tools repo locally, or GitHub raw if installed standalone)
| Standard File | Purpose |
|---|
typing.md | Type hint coverage, modern annotation syntax, static analysis |
code-style.md | PEP 8, naming conventions, import grouping, docstrings, Ruff |
function-design.md | Single responsibility, function length, side-effect isolation |
data-validation.md | Pydantic boundaries, validators, typed responses, settings |
error-handling.md | Exception specificity, custom hierarchy, logging levels |
module-structure.md | src/ layout, __all__, layer boundaries, circular imports |
testing-unit.md | Unit test patterns, fixtures, parametrize, mocking |
testing-integration.md | Integration tests, teardown, isolation, markers |
Report Metadata (MANDATORY)
Every generated report MUST include a metadata block at the very end. This is non-negotiable — never omit it.
To resolve the source and version:
- Look for
.claude-plugin/plugin.json by traversing up from this skill's directory
- If found, read
name and version from that file (plugin context)
- If not found, use
Somnio CLI as the name and unknown as the version (CLI context)
Include this block at the very end of the report:
---
Generated by: [plugin name or "Somnio CLI"] v[version]
Skill: python-best-practices
Date: [YYYY-MM-DD]
Somnio AI Tools: https://github.com/somnio-software/somnio-ai-tools
---