Skip to content

Repository files navigation

Werkstatt

What is Werkstatt?

Werkstatt began as a fork of Jesse Vincent's MIT-licensed Superpowers. We kept the core idea, a skill-based development workflow for coding agents, and continued tuning it for harder engineering work: stronger default model selection, stricter review gates, clearer platform mappings, and fork-owned installation paths.

Werkstatt is built on a set of composable "skills" and initial instructions that make sure your agent uses them.

When to use it

Werkstatt is built for high-stakes, complex work — the kind of session where you want the agent to slow down, brainstorm, plan, and review before it touches code. For small edits, simple bug fixes, or basic features, the full workflow is overkill and will feel like friction.

The harness is also token-hungry: every session loads skill content, runs review passes, and dispatches subagents. For lightweight sessions, you may want to deactivate it (disable the plugin, or skip the skill prefix) and re-enable it when you're back on something that warrants the discipline.

Installation

Note: Installation differs by platform. Claude Code uses the marketplace file in this repository. Codex and OpenCode require manual setup.

Claude Code (via Plugin Marketplace)

In Claude Code, register the marketplace first:

/plugin marketplace add Bollwerkio/werkstatt

Then install the plugin from this marketplace:

/plugin install werkstatt@werkstatt-marketplace

Cursor

The repository includes a Cursor plugin manifest, but Werkstatt is not currently listed in the public Cursor Marketplace. Once published there, it should be installable from Cursor Agent chat with:

/add-plugin werkstatt

Until then, use the Claude Code, Codex, or OpenCode installation paths.

Codex App

In the Codex app, open the Plugins sidebar, add this repository, and install Werkstatt:

https://github.com/Bollwerkio/werkstatt

Codex CLI

Tell Codex:

Fetch and follow instructions from https://raw.githubusercontent.com/Bollwerkio/werkstatt/refs/heads/main/.codex/INSTALL.md

Detailed docs: docs/README.codex.md

OpenCode

Tell OpenCode:

Fetch and follow instructions from https://raw.githubusercontent.com/Bollwerkio/werkstatt/refs/heads/main/.opencode/INSTALL.md

Detailed docs: docs/README.opencode.md

GitHub Copilot CLI

copilot plugin marketplace add Bollwerkio/werkstatt
copilot plugin install werkstatt@werkstatt-marketplace

Gemini CLI

Werkstatt includes Gemini CLI metadata for experimentation, but Gemini is not a suitable harness for serious Werkstatt coding sessions today. The skills depend on reliable multi-step development, review discipline, and exact tool mapping. Gemini does not meet that bar.

gemini extensions install https://github.com/Bollwerkio/werkstatt

To update:

gemini extensions update werkstatt

Verify Installation

Start a new session in your chosen platform and ask for something that should trigger a skill (for example, "help me plan this feature" or "let's debug this issue"). The agent should automatically invoke the relevant werkstatt skill.

How it works

It starts from the moment you fire up your coding agent. As soon as it sees that you're building something, it doesn't just jump into trying to write code. Instead, it steps back and asks you what you're really trying to do.

Once it's teased a spec out of the conversation, it shows it to you in chunks short enough to actually read and digest.

After you've signed off on the design, your agent puts together an implementation plan that's clear enough for an enthusiastic junior engineer with poor taste, no judgement, no project context, and an aversion to testing to follow. It emphasizes true red/green TDD, YAGNI (You Aren't Gonna Need It), and DRY.

Next up, once you say "go", it launches a subagent-driven-development process, having agents work through each engineering task, inspecting and reviewing their work, and continuing forward. It's not uncommon for Claude to be able to work autonomously for a couple hours at a time without deviating from the plan you put together.

There's a bunch more to it, but that's the core of the system. And because the skills trigger automatically, you don't need to do anything special. Your coding agent just has Werkstatt.

The Basic Workflow

  1. brainstorming - Activates before writing code. Refines rough ideas through questions, explores alternatives, presents design in sections for validation. Saves design document.

  2. using-git-worktrees - Activates after design approval. Creates isolated workspace on new branch, runs project setup, verifies clean test baseline.

  3. writing-plans - Activates with approved design. Breaks work into bite-sized tasks (2-5 minutes each). Every task has exact file paths, complete code, verification steps.

  4. subagent-driven-development or executing-plans - Activates with plan. Dispatches fresh subagent per task with two-stage review (spec compliance, then code quality), or executes in batches with human checkpoints.

  5. test-driven-development - Activates during implementation. Enforces RED-GREEN-REFACTOR: write failing test, watch it fail, write minimal code, watch it pass, commit. Deletes code written before tests.

  6. requesting-code-review - Activates between tasks. Reviews against plan, reports issues by severity. Critical issues block progress.

  7. finishing-a-development-branch - Activates when tasks complete. Verifies tests, presents options (merge/PR/keep/discard), cleans up worktree.

The agent checks for relevant skills before any task. Mandatory workflows, not suggestions.

What's Inside

Skills Library

Testing

  • test-driven-development - RED-GREEN-REFACTOR cycle (includes testing anti-patterns reference)

Debugging

  • systematic-debugging - 4-phase root cause process (includes root-cause-tracing, defense-in-depth, condition-based-waiting techniques)
  • verification-before-completion - Ensure it's actually fixed

Collaboration

  • brainstorming - Socratic design refinement
  • writing-plans - Detailed implementation plans
  • executing-plans - Batch execution with checkpoints
  • dispatching-parallel-agents - Concurrent subagent workflows
  • requesting-code-review - Pre-review checklist
  • receiving-code-review - Responding to feedback
  • using-git-worktrees - Parallel development branches
  • finishing-a-development-branch - Merge/PR decision workflow
  • subagent-driven-development - Fast iteration with two-stage review (spec compliance, then code quality)

Meta

  • writing-skills - Create new skills following best practices (includes testing methodology)
  • using-werkstatt - Introduction to the skills system

Philosophy

  • Test-Driven Development - Write tests first, always
  • Systematic over ad-hoc - Process over guessing
  • Complexity reduction - Simplicity as primary goal
  • Evidence over claims - Verify before declaring success

Werkstatt is forked from Superpowers by Jesse Vincent — read the original announcement: Superpowers for Claude Code.

Contributing

Skills live directly in this repository. To contribute:

  1. Fork the repository
  2. Create a branch for your skill
  3. Follow the writing-skills skill for creating and testing new skills
  4. Submit a PR

See skills/writing-skills/SKILL.md for the complete guide.

Updating

Skills update automatically when you update the plugin:

/plugin update werkstatt

License

MIT License - see LICENSE file for details

Community

Werkstatt is maintained by Bollwerk AI. It is a fork of Superpowers by Jesse Vincent and the team at Prime Radiant, to whom we owe the original work.

Support

About

Skills, hooks, and workflow discipline for Claude Code, Codex, Cursor & OpenCode. Turn an AI coding assistant into a rigorous engineer.

Resources

Code of conduct

Stars

7 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages