event-storming

event-storming

A Claude Code plugin for EventStorming, Alberto Brandolini's workshop format for collaborative domain discovery. It helps you explore a business domain, discover bounded contexts, and bridge the resulting model to DDD tactical design.

It ships two skills, split by whether you want to read facilitation guidance or run an agentic workshop:

Skill	Invoke	Does
methodology	/event-storming:methodology [--big-picture|--process|--design-level|--patterns|--glossary|--notation|--remote]	Facilitation knowledge and reference across the three formats: notation, building blocks, patterns/anti-patterns, remote adaptations. No args runs an interactive discovery flow.
simulation	/event-storming:simulation [--simulate|--process-model|--design-level|--evaluate|--retrospective|--induction|--value|--crc|--ux|--discover-bcs] [domain]	An agentic, multi-persona EventStorming workshop that produces a structured-markdown model by default, plus per-run scoring and bounded-context discovery. A live Miro-board rendering path is available when the first-party miro plugin is enabled (see Requirements).

When to use which

• methodology is pure reference. You are facilitating or learning EventStorming yourself and want the format guidance, notation, and patterns at hand. It needs no external tools.
• simulation runs the workshop for you: it spins up 4 to 7 siloed personas, plays every Big Picture phase in order, discovers bounded contexts, and lets you drill into Process Modeling and Design-Level per context. By default it emits the model as structured markdown (event timeline, persona roster, bounded-context tables); with the first-party miro plugin enabled it renders the same model onto a live Miro board instead.

Each skill auto-invokes on its own trigger phrases, or you can call it directly.

Install

/plugin marketplace add melodic-software/claude-code-plugins
/plugin install event-storming@melodic-software

Requirements

• methodology. None beyond Claude Code.
• simulation. No hard dependency. It runs out of the box in structured-markdown mode (the agentic workshop, emitting the model as an event timeline, persona roster, and bounded-context tables). Two optional surfaces enhance it, neither bundled:
  • The first-party miro plugin for the live-board rendering path.
  • A web-research surface for domain context.

Both optional surfaces degrade gracefully:

• Miro. The live-board path uses the first-party miro plugin: a bundled local-stdio MCP server that a fresh consumer must install from an available marketplace and then enable (select the miro plugin in /plugin, then enable it) with a Miro API token, entered when Claude Code asks for it. Its tools are namespaced mcp__plugin_miro_miro__*, or mcp__miro__* under the miro README's user-scope override; the gates accept either. With the plugin absent (or its tools unavailable), simulation detects it at preflight and runs in structured-markdown mode instead of failing. Miro's official hosted server (mcp.miro.com) was evaluated and rejected as the target (no board-delete tool for teardown; third-party remote egress). Trust record in the marketplace repo's MCP decision table (docs/migration-playbook.md).
• Web research absent. The skills use the Perplexity MCP tools if present, otherwise Claude Code's built-in WebSearch / WebFetch; with no research surface at all they ask you for the domain context rather than guessing.

Shareable exports (slides/docs/PDF) hand off to the document-skills plugin when it is installed, and fall back to markdown otherwise.

Configuration

This plugin has no userConfig. Its inputs are conversational (the domain and mode you pass) plus the optional MCP servers above. Generated state (session archives, run history, comparison baselines) is written to the plugin's own persistent data directory (${CLAUDE_PLUGIN_DATA}), never into your project tree.

License

MIT (SPDX-License-Identifier: MIT).