dumbdumbdumb

dumbdumbdumb is a single Python file analyzer running multiple quality tools, outputting a single normalized JSON report.

What it does

Given one target file, the analyzer runs:

• ruff (fast linting + common correctness checks)
• black --check (format compliance)
• mypy --strict (strict static type checking)
• pylint (deeper style/design/code-smell analysis)
• flake8 (PEP8 and plugin-based lint checks)
• complexipy (reports cognitive complexity per file/function)

Quick start

1) Build the Docker image

./build.sh

This tags localhost/dumbdumbdumb:latest (the localhost/ prefix keeps Podman from asking which registry to use, which fails when stdout is not a TTY—for example ./dumbdumbdumb file.py > out.json). Re-run it whenever you pull changes or want refreshed analyzer versions inside the image.

2) Run against a file

You only need Docker or Podman and bash. The wrapper script is self-contained: copy dumbdumbdumb anywhere on your PATH (for example ~/bin/) if you like; it always runs localhost/dumbdumbdumb:latest.

chmod +x dumbdumbdumb   # once, if needed
./dumbdumbdumb path/to/file.py

Pretty JSON:

./dumbdumbdumb --pretty path/to/file.py

To work on the in-container Python code, use a normal editable install (python -m pip install -e .) while iterating; production use is still Docker + this script.

JSON report contract

Top-level fields:

• target: analyzed file path as seen inside the container
• generated_at: UTC ISO-8601 timestamp
• summary: aggregate counts and recommended process exit code
• findings: flat list of all normalized findings from every tool (in run order)
• tools: per-tool execution summary (no raw stdout/stderr; use the tool directly if you need that)

summary includes:

• tool_count
• finding_count (all severities)
• actionable_finding_count (excludes severity info, e.g. complexity notes)
• suppressed_finding_count (aggregated across tools)
• failed_tools (names where execution failed)
• tools_with_issues (names that reported lint/type issues)
• exit_code: 0 clean, 1 actionable findings or tools with issues, 2 at least one tool failed to run

Each object in findings includes:

• tool, severity, message, path
• Optional: line, column, end_line, end_column, code, symbol (omitted when not applicable)

Each object in tools includes:

• name, status (ok | issues | error), exit_code, duration_ms
• finding_count, suppressed_finding_count
• error (present only when status is error)

Docker image notes

• Using Docker python-alpine instead of python-slim will result in 1GB bigger image, since it has to install build-base