Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
Mindaugas Morkunas avatar

Ai Workshop

  • 3 repo stars
  • Updated May 20, 2026
  • Visma-Tech-LT/ai-workshop-marketplace

ai-workshop is a Claude Code plugin marketplace that provides spec-driven workflow skills and agents so developers can research, plan, implement, and review code in deliberate phases.

About

ai-workshop is a Claude Code plugin marketplace for Visma's AI Engineering Workshop Part 2 spec-driven workflow. It bundles six agents and five skills that walk you from /sdw:research-codebase through planning, implementation, and /sdw:code-review with explicit human checkpoints between phases. Use it when you want disciplined agent-assisted delivery where markdown artifacts carry context instead of long chat threads.

  • Six specialized agents for locating, analyzing, and reviewing code
  • Five /sdw workflow skills from research through code review
  • Outputs research.md, plan.md, and review.md under .claude/ai/
  • Intentional /clear between phases to use documents as contracts
  • Optional skeptic gate to adversarially challenge plans

Ai Workshop by the numbers

  • Data as of Jul 7, 2026 (Skillselion catalog sync)
/plugin marketplace add Visma-Tech-LT/ai-workshop-marketplace

Add your badge

Show developers this marketplace is listed on Skillselion. Paste this into your README.

Listed on Skillselion
repo stars3
Last updatedMay 20, 2026
RepositoryVisma-Tech-LT/ai-workshop-marketplace

How do I run a spec-driven agent workflow where each phase produces reviewable documents instead of losing context in one long chat?

Run a manual spec-driven workflow from codebase research through planning, implementation, and code review with dedicated agents.

Who is it for?

Developers learning or practicing spec-driven development who will pause to review research.md and plan.md between phases.

Skip if: Teams seeking a fully autonomous one-shot codegen pipeline without human checkpoints between steps.

What you get

Structured markdown artifacts and agent-assisted code changes accumulate under .claude/ai/ with a final review.md from code review.

Recommended Marketplaces

FAQ

What is the recommended workflow order?

The README defines /sdw:research-codebase, /sdw:create-plan, optional skeptic review, /sdw:implement-plan, then /sdw:code-review.

Which agents are included?

Agents listed are codebase-locator, codebase-analyzer, codebase-pattern-finder, test-runner, code-reviewer, and web-search-researcher.

Why does the workflow require /clear between phases?

The README states documents like research.md and plan.md are the handoff contract, forcing review instead of relying on accumulated chat context.

AI Agentsagentsautomation

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.