Most claims about "AI for architecture governance" don't hold up beyond the first template.

It's easy to generate a few document outlines, but actually stringing together principles, stakeholders, requirements, design reviews, and a traceability matrix — with tangible outputs produced at every step — is an entirely different story.

ArcKit builds this entire workflow directly into Claude Code.

Instead of just giving you a prompt for writing documents, it uses slash commands to drive the full lifecycle:

- Start with defining principles and stakeholders

- Move on to requirements decomposition

- Conduct structured design reviews

- Generate a complete traceability matrix

Every step outputs a tangible markdown artifact. Token costs are clearly stated upfront in the readme, with no hidden fine print.

Three-step installation:

```

/plugin marketplace

```

Add `tractorjuice/arc-kit`, then run:

```

claude plugin install arckit

```

One user commented: "Only when reviews and traceability are formalized as Markdown artifacts do they actually become part of the engineering process."

That hits the nail on the head. In AI toolchains, integrating into actual workflows and producing inspectable intermediate artifacts is far more valuable than just "generating a pretty architecture document".


ArcKit's positioning is "building better enterprise architecture through structured strategy, design, delivery, and assurance workflows". Its core engine covers modules including requirements management, decision records, risk registers, and data models, and runs leveraging Claude Code's observer and overlay mechanisms.

The value of this type of tool doesn't lie in "AI's ability to draw architecture diagrams" — it lies in turning governance processes into executable, traceable, auditable engineering practices.

Project repository: github.com/tractorjuice/arc-kit