Reference¶

Accurate technical description of the tools. Aimed at users who need facts.

Code review tools¶

Scanners, review conventions, and language-specific style rules.

Document	Description
Scripts reference	All scanning scripts: usage, options, examples
Review mode	Read-only review conventions
Code exploration	tokensave decision tree and task templates
tokensave	tokensave MCP tools reference
Python code style	Python style rules and configurable toggles

Drift detection, staleness scoring, and operational conventions.

Document	Description
Drift detection tools	drift_analyzer, staleness scorer, API validator, link checker
Workflows & tools	Quick start, 5 core workflows, CI recipes
Scoring & categories	Staleness scoring model, drift categories, troubleshooting
Operational rules	Tool invocation conventions and drift report format

Settings, subcommands, CI/CD, frontmatter, reports, and tool messages.

Document	Description
Settings schema	Full field-by-field reference for `settings.json`
Subcommands	All Zolletta-metaskill subcommands
CI/CD workflows	GitHub Actions workflows: scope and triggers
Frontmatter	SKILL.md frontmatter fields
Reports	Report file format and templates
Tool messages	"Not installed" messages for the tool-failure handler

A real review output from dogfooding — see what the reports look like.

Document	Description
Overview	Real review output from dogfooding — overview and files
SUMMARY.md	Executive summary with overall grade and trend
TODO.md	Prioritized action items (critical, high, medium, low)
patterns.md	Design pattern review — God classes, SOLID, coupling
documentor.md	Documentation review — Diátaxis compliance, drift
python-code-style.md	Python source code style — ruff, mypy, naming, docstrings
python-testing-style.md	Python test code review — isolation, naming, coverage