Awesome Reviewers

Verifying Python changes

Files: .py, .pyi. Derived from 64 scanned source repositories and 707 review prompts on https://awesomereviewers.com. Page: https://awesomereviewers.com/verify/python/

How to use this: run every check under Run on every change. Then, for each section under Add when the change touches that describes your diff, run its checks too. Checks catch what a tool can find; then apply the listed review prompts for what tools miss (fetch each prompt’s raw URL).

Usage is the share of scanned repositories in this language that run the tool. A missing tool means no scanned repository runs one the scanner recognises: fall back to the prompts.

Run on every change

• Formatting: ruff-format (50% of repos). Command seen: ruff format .. Autofixes. Exit codes: 0 already formatted; 1 files would be reformatted (–check); 2 error. Alternatives: black (14%), isort (12%). Docs: https://docs.astral.sh/ruff/formatter/#exit-codes
• Linting: ruff-check (72% of repos). Command seen: ruff check .. Autofixes. Exit codes: 0 no violations; 1 violations found; 2 invalid configuration, CLI options or internal error. Alternatives: flake8 (20%), pylint (12%). Docs: https://docs.astral.sh/ruff/linter/#exit-codes
• Type checking: mypy (42% of repos). Command seen: mypy. Exit codes: 0 no type errors; 1 type errors found; 2 crash or invalid invocation. Alternatives: pyright (17%). Docs: https://mypy.readthedocs.io/en/stable/command_line.html
• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html

Review prompts (111 in total, most discussed first):

• Use descriptive names — strands-agents/sdk-python, Naming Conventions
• Consistent module organization — p-e-w/heretic, Code Style
• Avoid code duplication — BerriAI/litellm, Code Style
• eliminate unnecessary code — Unstructured-IO/unstructured, Code Style
• Readable formatting and idioms — celery/celery, Code Style
• eliminate code duplication — sgl-project/sglang, Code Style
• Use semantically clear names — Unstructured-IO/unstructured, Naming Conventions
• Eliminate redundant constructs — apache/mxnet, Code Style
• Centralize shared utilities — awslabs/mcp, Code Style
• Collision-Safe Identifier Mapping — nousresearch/hermes-agent, Naming Conventions
• Idiomatic, minimal Python constructs — pallets/flask, Code Style
• Use clear names — p-e-w/heretic, Naming Conventions

Add when the change touches

Authentication, input handling, secrets, permissions

• Security scanning: codeql (28% of repos). Runs via GitHub Action github/codeql-action/analyze. Alternatives: bandit (9%), semgrep (8%). Docs: https://github.com/github/codeql
• Secret scanning: gitleaks (2% of repos). Runs via pre-commit hook gitleaks. Docs: https://github.com/gitleaks/gitleaks
• Dependency audit: dependency-review (8% of repos). Runs via GitHub Action actions/dependency-review-action. Alternatives: pip-audit (3%). Docs: https://github.com/actions/dependency-review-action

Review prompts (48 in total, most discussed first):

• Prevent sensitive data exposure — bridgecrewio/checkov, Security
• Protect sensitive data — bridgecrewio/checkov, Security
• Boundary Hardening Rules — nousresearch/hermes-agent, Security
• Fail-Closed Input Sanitization — awslabs/mcp, Security
• Secure Untrusted Boundaries — Azure/azure-cli, Security
• Trust Boundaries And Safe Parsing — the-pr-agent/pr-agent, Security
• Secure Auth and Logging — openai/openai-python, Security
• Sanitize Untrusted Boundaries — mvanhorn/last30days-skill, Security
• Centralized input guards — infiniflow/ragflow, Security
• Validate user-controlled paths — apache/airflow, Security
• Least privilege principle — prowler-cloud/prowler, Security
• Defensive Security Validation — openai/codex-security, Security

Public APIs, signatures, error paths, nullability

• Type checking: mypy (42% of repos). Command seen: mypy. Exit codes: 0 no type errors; 1 type errors found; 2 crash or invalid invocation. Alternatives: pyright (17%). Docs: https://mypy.readthedocs.io/en/stable/command_line.html
• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html
• Property-based tests: hypothesis (14% of repos). Docs: https://github.com/HypothesisWorks/hypothesis

Review prompts (167 in total, most discussed first):

• Actionable Error Propagation — Azure/azure-cli, Error Handling
• Avoid silent error masking — awslabs/mcp, Error Handling
• Exception Handling Policy — the-pr-agent/pr-agent, Error Handling
• Validate Inputs Explicitly — TheAlgorithms/Python, Error Handling
• Surface errors clearly — p-e-w/heretic, Error Handling
• explicit None handling — Unstructured-IO/unstructured, Null Handling
• Deterministic exception handling — celery/celery, Error Handling
• Safe null handling — langgenius/dify, Null Handling
• Null-safe Defaults — the-pr-agent/pr-agent, Null Handling
• Use Null-Safe Config Reads — nousresearch/hermes-agent, Null Handling
• Accurate API Contracts — awslabs/mcp, API
• Standardize Typed Errors — openai/openai-python, Error Handling

Service and framework code

• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html
• Version and platform matrix: ci-matrix (56% of repos). Command seen: analyze (include: 2). Docs: https://docs.github.com/en/actions/how-tos/write-workflows/choose-what-workflows-do/run-job-variations

Review prompts (6 in total, most discussed first):

• Test Canvas Edge Cases — celery/celery, Celery
• Ensure deterministic resume behavior — strands-agents/sdk-python, Temporal
• Offload heavy operations — langgenius/dify, Celery
• leverage framework capabilities — PostHog/posthog, Temporal
• Task signature methods — prowler-cloud/prowler, Celery
• Use acks_late for reliability — getsentry/sentry, Celery

Algorithms, parsers, data transforms

• Property-based tests: hypothesis (14% of repos). Docs: https://github.com/HypothesisWorks/hypothesis
• Fuzzing — no tool detected in the sample. Catches: Crashes, panics and hangs on malformed input, especially in parsers and decoders.
• Mutation testing: mutmut (2% of repos). Command seen: mutmut run 2>&1. Docs: https://github.com/boxed/mutmut
• Proofs and specifications: lean (2% of repos). Docs: https://github.com/leanprover/lean4

Review prompts (47 in total, most discussed first):

• Choose optimal algorithms — bridgecrewio/checkov, Algorithms
• Deterministic Algorithm Correctness — langchain-ai/langchain, Algorithms
• optimize algorithmic efficiency — python-poetry/poetry, Algorithms
• Prevent algorithmic pitfalls — boto/boto3, Algorithms
• Precise Matching Rules — nousresearch/hermes-agent, Algorithms
• Algorithm Correctness Standards — TheAlgorithms/Python, Algorithms
• implement algorithms correctly — stanfordnlp/dspy, Algorithms
• Optimize iteration patterns — apache/mxnet, Algorithms
• Verify algorithmic equivalence — python/cpython, Algorithms
• Consistent, Guarded Ranking — mvanhorn/last30days-skill, Algorithms
• Deterministic Mapping Pipeline — infiniflow/ragflow, Algorithms
• Optimization algorithm compatibility — p-e-w/heretic, Algorithms

Threads, async code, locks, shared state

• Concurrency model checking — no tool detected in the sample. Catches: Races and orderings that fail only under specific interleavings.
• Undefined-behaviour detection — no tool detected in the sample. Catches: Memory-safety and aliasing violations in unsafe code.
• Property-based tests: hypothesis (14% of repos). Docs: https://github.com/HypothesisWorks/hypothesis

Review prompts (30 in total, most discussed first):

• Nonblocking Concurrency Rules — nousresearch/hermes-agent, Concurrency
• Avoid shared mutable state — langchain-ai/langchain, Concurrency
• Thread-safe lifecycle isolation — celery/celery, Concurrency
• Safe Async Resource Handling — awslabs/mcp, Concurrency
• Avoid event-loop blocking — openai/openai-python, Concurrency
• Choose appropriate synchronization — strands-agents/sdk-python, Concurrency
• Ensure operation completion safety — LMCache/LMCache, Concurrency
• Non-Blocking Async Recovery — infiniflow/ragflow, Concurrency
• Concurrent resource management — langgenius/dify, Concurrency
• Deterministic concurrency tests — python/cpython, Concurrency
• use context managers concurrency — commaai/openpilot, Concurrency
• background task coordination — BerriAI/litellm, Concurrency

Network calls, retries, timeouts, protocols

• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html
• Fuzzing — no tool detected in the sample. Catches: Crashes, panics and hangs on malformed input, especially in parsers and decoders.

Review prompts (10 in total, most discussed first):

• Restrict public access — bridgecrewio/checkov, Networking
• Restrict public network — bridgecrewio/checkov, Networking
• Configure HTTP requests properly — python-poetry/poetry, Networking
• Preserve protocol semantics — python/cpython, Networking
• Robust aiohttp Networking — openai/openai-python, Networking
• HTTP protocol correctness — pallets/flask, Networking
• Manage network resources — anthropics/anthropic-sdk-python, Networking
• network service identification — LMCache/LMCache, Networking
• validate connection timeouts — browser-use/browser-use, Networking
• Platform-specific socket configuration — commaai/openpilot, Networking

UI components and pages

• End-to-end browser tests — no tool detected in the sample. Catches: User flows broken across components, routing and the network.
• Linting: ruff-check (72% of repos). Command seen: ruff check .. Autofixes. Exit codes: 0 no violations; 1 violations found; 2 invalid configuration, CLI options or internal error. Alternatives: flake8 (20%), pylint (12%). Docs: https://docs.astral.sh/ruff/linter/#exit-codes

Databases, migrations, caching

• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html
• Version and platform matrix: ci-matrix (56% of repos). Command seen: analyze (include: 2). Docs: https://docs.github.com/en/actions/how-tos/write-workflows/choose-what-workflows-do/run-job-variations

Review prompts (34 in total, most discussed first):

• SQLAlchemy 2.0 patterns — langgenius/dify, Database
• Cache Scope and Keys — nousresearch/hermes-agent, Caching
• Safe caching practices — langchain-ai/langchain, Caching
• Credential-Aware Caching — awslabs/mcp, Caching
• Cache invalidation consistency — PostHog/posthog, Caching
• Use connection by alias — django/django, Database
• Cache validity and bounds — openai/openai-python, Caching
• Cache state consistency — LMCache/LMCache, Caching
• Prevent N+1 database queries — getsentry/sentry, Database
• Cache Only After Inputs — celery/celery, Caching
• Optimize cache performance — sgl-project/sglang, Caching
• Split complex migrations incrementally — PostHog/posthog, Migrations

Hot paths and performance

• No tool-based check applies; rely on the prompts.

Review prompts (36 in total, most discussed first):

• Avoid Unnecessary Work — celery/celery, Performance Optimization
• Bounded Hot-Path Work — nousresearch/hermes-agent, Performance Optimization
• Bound Work and Queries — infiniflow/ragflow, Performance Optimization
• Manage tensors proactively — p-e-w/heretic, Performance Optimization
• Eliminate redundant operations — sgl-project/sglang, Performance Optimization
• Vectorize over Python loops — vllm-project/vllm, Performance Optimization
• Budget-Aware Performance — the-pr-agent/pr-agent, Performance Optimization
• Enforce Resource Caps — awslabs/mcp, Performance Optimization
• Performance-Sensitive Efficiency — TheAlgorithms/Python, Performance Optimization
• Optimize hot-path code — python/cpython, Performance Optimization
• Avoid Hidden Performance Traps — langchain-ai/langchain, Performance Optimization
• Cache expensive operations — PostHog/posthog, Performance Optimization

Model, prompt and LLM code

• Unit and integration tests: pytest (86% of repos). Command seen: pytest. Exit codes: 0 all tests passed; 1 some tests failed; 2 interrupted by the user; 3 internal error; 4 command line usage error; 5 no tests were collected. Alternatives: unittest (12%). Docs: https://docs.pytest.org/en/stable/reference/exit-codes.html

Review prompts (40 in total, most discussed first):

• Model integration rules — p-e-w/heretic, AI
• prefer OpenAI compatibility — BerriAI/litellm, AI
• Improve code organization — sgl-project/sglang, AI
• Device-agnostic acceleration code — pytorch/pytorch, AI
• Avoid magic numbers — vllm-project/vllm, AI
• Fail-closed evaluation metrics — promptfoo/promptfoo, AI
• Model-Safe Parameter Validation — the-pr-agent/pr-agent, AI
• Preserve AI Structure Invariants — langchain-ai/langchain, AI
• Handle optional dependencies gracefully — stanfordnlp/dspy, AI
• leverage existing framework functionality — ggml-org/llama.cpp, AI
• LLM Interface Contracts — awslabs/mcp, AI
• preserve dtype and shapes — p-e-w/heretic, Pytorch

Logging, metrics, tracing

• No tool-based check applies; rely on the prompts.

Review prompts (26 in total, most discussed first):

• Scoped, Accurate Warning Logs — Azure/azure-cli, Logging
• Logging Correctness Rules — celery/celery, Logging
• Follow logging best practices — vllm-project/vllm, Logging
• Logging consistency and safety — the-pr-agent/pr-agent, Logging
• Use appropriate log levels — langgenius/dify, Logging
• Use appropriate logging levels — home-assistant/core, Logging
• Use proper logging mechanisms — BerriAI/litellm, Logging
• Structure logs with context — getsentry/sentry, Logging
• Sanitize observability data exports — open-telemetry/opentelemetry-python, Observability
• Add logging without duplication — docker/compose, Logging
• Structure logs effectively — django/django, Logging
• Use appropriate logging levels — bridgecrewio/checkov, Logging

CI workflows and configuration files

• CI workflow linting — no tool detected in the sample. Catches: Broken GitHub Actions syntax, script injection, over-broad token permissions.
• File hygiene: pre-commit-hooks (39% of repos). Runs via pre-commit hook end-of-file-fixer. Autofixes. Docs: https://github.com/pre-commit/pre-commit-hooks

Review prompts (68 in total, most discussed first):

• Configuration ownership and determinism — p-e-w/heretic, Configurations
• Centralized Config Standards — eosphoros-ai/DB-GPT, Configurations
• Centralize environment variables — bridgecrewio/checkov, Configurations
• validate configuration defaults — Unstructured-IO/unstructured, Configurations
• Use configurable default values — prowler-cloud/prowler, Configurations
• Explicit config behavior — celery/celery, Configurations
• Robust environment config handling — pallets/flask, Configurations
• Language-agnostic configuration design — oraios/serena, Configurations
• Config.yaml First Policy — nousresearch/hermes-agent, Configurations
• Region configuration consistency — awslabs/mcp, Configurations
• Optional configuration parameters — stanfordnlp/dspy, Configurations
• Feature flag rollouts — getsentry/sentry, Configurations

Dependencies and lockfiles

• Dependency audit: dependency-review (8% of repos). Runs via GitHub Action actions/dependency-review-action. Alternatives: pip-audit (3%). Docs: https://github.com/actions/dependency-review-action
• Lockfile consistency: uv-lock (20% of repos). Command seen: uv lock --check. Autofixes. Docs: https://github.com/astral-sh/uv
• Version and platform matrix: ci-matrix (56% of repos). Command seen: analyze (include: 2). Docs: https://docs.github.com/en/actions/how-tos/write-workflows/choose-what-workflows-do/run-job-variations

The tests themselves

• Mutation testing: mutmut (2% of repos). Command seen: mutmut run 2>&1. Docs: https://github.com/boxed/mutmut
• Version and platform matrix: ci-matrix (56% of repos). Command seen: analyze (include: 2). Docs: https://docs.github.com/en/actions/how-tos/write-workflows/choose-what-workflows-do/run-job-variations

Review prompts (52 in total, most discussed first):

• Thorough edge-case verification — python/cpython, Testing
• Mock tests in tests/litellm — BerriAI/litellm, Testing
• Parameterize similar tests — python-poetry/poetry, Testing
• Mock external dependencies only — home-assistant/core, Testing
• Automate All Behavior Tests — TheAlgorithms/Python, Testing
• Behavioral Regression Testing — nousresearch/hermes-agent, Testing
• Prefer Robust Testing Assertions — langchain-ai/langchain, Testing
• Write clear test cases — Unstructured-IO/unstructured, Testing
• Use idiomatic pytest features — pallets/flask, Testing
• Deterministic recorded testing — Azure/azure-cli, Testing
• Meaningful Tests For Guards — celery/celery, Testing
• Write self-documenting tests — boto/boto3, Testing

Documentation and comments

• Docs build: sphinx (12% of repos). Command seen: sphinx-build -j2 -b configcheck -d {envtmpdir}/doctrees docs docs/_build/configcheck. Alternatives: mkdocs (6%). Docs: https://github.com/sphinx-doc/sphinx
• Spelling: codespell (25% of repos). Command seen: codespell --toml pyproject.toml --write. Docs: https://github.com/codespell-project/codespell

Review prompts (32 in total, most discussed first):

• Docstring Standards — TheAlgorithms/Python, Documentation
• Docstrings With Intent — langchain-ai/langchain, Documentation
• PEP 257 docstring compliance — Unstructured-IO/unstructured, Documentation
• Document code thoroughly — vllm-project/vllm, Documentation
• Concise, Correct CLI Help — Azure/azure-cli, Documentation
• Document Behavior Changes — celery/celery, Documentation
• Document Behavior Accurately — awslabs/mcp, Documentation
• Complete parameter documentation — stanfordnlp/dspy, Documentation
• Document code rationale — pydantic/pydantic, Documentation
• Write purposeful comments — open-telemetry/opentelemetry-python, Documentation
• Comprehensive API documentation — boto/boto3, Documentation
• Accurate, synced documentation — python/cpython, Documentation

Source repositories

nousresearch/hermes-agent, TheAlgorithms/Python, tensorflow/tensorflow, langchain-ai/langchain, langgenius/dify, langflow-ai/langflow, pytorch/pytorch, fastapi/fastapi, django/django, comfyanonymous/ComfyUI, home-assistant/core, infiniflow/ragflow, python/cpython, microsoft/markitdown, pallets/flask, TauricResearch/TradingAgents, browser-use/browser-use, commaai/openpilot, mvanhorn/last30days-skill, karpathy/nanochat, vllm-project/vllm, calesthio/OpenMontage, getsentry/sentry, apache/airflow, docker/compose, Aider-AI/aider, pola-rs/polars, crewaiinc/crewai, python-poetry/poetry, openai/openai-python, celery/celery, PostHog/posthog, BerriAI/litellm, stanfordnlp/dspy, pydantic/pydantic, apache/mxnet, eosphoros-ai/DB-GPT, agent0ai/agent-zero, sgl-project/sglang, SWE-agent/SWE-agent, bytedance/deer-flow, oraios/serena, semgrep/semgrep, the-pr-agent/pr-agent, emcie-co/parlant, Unstructured-IO/unstructured, prowler-cloud/prowler, huggingface/tokenizers, awslabs/mcp, boto/boto3, openai/codex-security, bytedance/trae-agent, openai/skills, bridgecrewio/checkov, p-e-w/heretic, Azure/azure-cli, strands-agents/sdk-python, LMCache/LMCache, anthropics/anthropic-sdk-python, awslabs/agentcore-samples, smallcloudai/refact, anthropics/claude-agent-sdk-python, salesforce/cloudsplaining, open-telemetry/opentelemetry-python

This site is open source. Improve this page.