Dispatch MTK setup workflows to bootstrap or audit repositories.
npx skills add https://github.com/moberghr/mtk-agent-toolkit --skill mtk-setup
Please help me install this Agent Skill. Skill: mtk-setup Source: https://github.com/moberghr/mtk-agent-toolkit/tree/main/.claude/skills/mtk-setup Command: npx skills add https://github.com/moberghr/mtk-agent-toolkit --skill mtk-setup
MTK-setup provides a single entry point to bootstrap a repo and re-audit its architecture, ensuring consistent onboarding and governance across projects.
Run /mtk-setup with no flags for a full bootstrap, or add --audit to re-run the architecture audit, or --merge to unify audits across repos.
High-intent search queries and answers about installing and using this skill.
To bootstrap a repository, you run the unified launcher with no flags. This dispatches to the first-time onboarding workflow, which generates standardized artifacts like CLAUDE.md, .claude/tech-stack references, and architecture principles for the project.
The best way to re-audit repository architecture is to run the launcher with the --audit flag. This dispatches to the architecture-audit refresh workflow, which evaluates the updated codebase and regenerates standardized architectural principles and references.
Yes, you can unify architecture audits across multiple MTK-enabled repositories by running the launcher with the --merge flag. This dispatches to the multi-repo merge workflow, which orchestrates and consolidates audits across the specified projects.
No external dependencies are required to use the MTK bootstrap and audit launcher. The Skill operates independently by implementing its own argument parsing, skill selection, and artifact generation to produce the necessary .claude references.
You should use the audit refresh workflow instead of a full bootstrap when an existing repository undergoes significant changes. A full bootstrap is intended only for first-time onboarding, while the audit refresh updates architectural principles without recreating the initial setup.
Argument parsing works by intercepting input flags to route the unified setup and audit orchestration to the correct workflow. Passing no flags triggers a full bootstrap, while --audit and --merge specifically select architecture re-audits and multi-repo consolidations respectively.
az9713Initialize a project with dark code prevention directories, CLAUDE.md section, and baseline audit.
hummbl-devAutomate GitHub repository initialization, governance, CI/CD, and documentation setup.
moberghrDetect tech stacks, audit codebases, and generate CLAUDE.md for MTK repositories.
recallnetBootstrap or audit canonical repo docs with governance presets.
