Mural Sticky Poster

A CLI tool for product teams (or anyone) to post stickies to a MURAL board from the command line. Useful for quickly populating boards with structured data. An included example is a process 'Event Storm'.

While the nature of event storming and many other visualisation techniques is collaborative, this tool offers a quick start, aiming to get something that someone has worked on in silo, like a text doc into a visual tool so the collaborative discussion can take centre stage.

Setup

Set up a MURAL app

1. Go to https://developers.mural.co and sign in
2. Navigate to Your apps → Create new app
3. Give it a name (e.g. "Sticky Poster CLI")
4. Set the redirect URI to exactly: http://127.0.0.1:8000/
5. Under scopes, enable murals:read and murals:write
6. Save the app
7. Copy the Client ID and Client Secret somewhere so you can add into your .env file later

Setup the python environment

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Configure credentials

Run the interactive setup:

python setup.py

It will prompt for:

• Client ID and Client Secret from your MURAL app
• Mural board URL or ID — paste the full board URL (e.g. https://app.mural.co/t/myworkspace/m/myworkspace/myworkspace.1780064070379/...) and it will extract the ID automatically, or enter the ID directly (e.g. myworkspace.1780064070379)

Config is saved to .env. Setup runs automatically on first use if .env is missing.

Usage

python main.py

On first run, a browser window opens for MURAL OAuth login.

You can then return to the CLI and you will be prompted to choose:

1. Post a single sticky — specify text, position, and hex colour
2. Post an event storm — from a JSON file (or the included example)
3. Generate event storm JSON — prints an LLM prompt you can use to convert a process doc into valid JSON for this tool
4. Re-run setup — update credentials or board ID

Post a single sticky

Ronseal - designed to make sure this thing is doing the thing it's supposed to do

Event Storm mode

Event storming is a colaborative process to explore a particular domain. Useful in software design and business modelling. The original intention for this tool was to help take some process docs that are currently written up in a variety of places in Confluence and map it in Mural to allow for a more collaborative conversation about the process we were mapping. Easing the initial setup of the workshop.

When posting an event storm, stickies are automatically laid out an order based on their type:

Sticky types

Type	Colour	Use for
event	Orange	Things that happened — "Config submitted"
command	Blue	Actions someone took — "Submit config"
actor	Yellow	Who did it — "Onboarding", "RP"
readmodel	Green	Info needed to make a decision — "Checklist"
policy	Purple	Business rules / automated reactions
system	Pink	Tools involved — "ServiceNow", "Slack"
hotspot	Red	Problems / pain points — "This takes 5 days"

Layout

Row 1 (y=0)    │ actor    │ actor    │ actor    │
Row 2 (y=200)  │ readmodel│ command  │ event    │ readmodel│ command  │ event    │ ...
Row 3 (y=400)  │ policy   │ policy   │
Row 4 (y=600)  │ system   │ system   │

• readmodel, command, event share the same row and flow left-to-right as the main timeline
• actor, policy, system each have their own row and flow independently
• hotspot is placed ad hoc with explicit x and y near the related sticky

JSON Schema

[
  {"type": "actor", "text": "Engagement Manager"},
  {"type": "readmodel", "text": "Onboarding checklist"},
  {"type": "command", "text": "Send onboarding manual"},
  {"type": "event", "text": "Manual sent to service team"},
  {"type": "policy", "text": "If incomplete - request more info"},
  {"type": "system", "text": "Google Docs"},
  {"type": "hotspot", "text": "This takes 5 days", "x": 500, "y": 300}
]

• Items are placed left-to-right in the order they appear
• Use "column": N to explicitly position a sticky at a specific column
• Hotspots require "x" and "y" fields

Generating Event Storm JSON from a document

Don't have your process in JSON yet? Option 3 in the CLI prints a ready-made prompt you can paste into any LLM (ChatGPT, Claude, etc.) along with your process document. The LLM will return valid JSON that works directly with this tool.

The prompt is also available at prompts/doc_to_event_storm.md if you want to use it outside the CLI.

Other modes

The tool is designed to be extensible to other types of visualisations. What's been implemented and some ideas for future iteration can be found in project.md