A map for unsolved matters

matters.global

matters.global is the map you build so your AI agent can actively move you toward your goals, with clear conditions, dependencies, and a visible path forward. It is the dependency layer between human intention and AI execution and enables AI to focus on what can actually move the world forward.

matters.global is software that turns human goals into an execution-ready dependency map.

It captures goals, blockers, and conditions, then maps their dependencies so you can see exactly what must happen before progress is possible.

Humans define intent and outcomes; AI agents help extract structure, recommend the next steps, and execute actions. For example, in science they can research the next paper, while in business they can draft a progress report.

01

Humans define the intent

You keep ownership of goals, risks, and what success should look like.

02

AI models the graph

Agents extract matters, conditions, and dependencies from your notes and prompts.

03

Humans choose actions

You decide the next step and what to keep private, public, or ask the team for.

Built for people blocked by unclear next steps.

01

Founders and product teams

Turn backlogs and launch questions into a dependency map with clear, executable next moves.

02

Researchers and operators

Turn notes and risks into a structured graph: what blocks progress, and what can move now.

03

Advisors and collaborators

Share a clean public view with partners without exposing private context.

From a messy situation into a useful map.

01

Name the matter

Capture the thing that matters enough to revisit: a goal, concern, risk, decision, or question.

02

Define resolution

Write down the observable conditions that would make it count as resolved.

03

Connect dependencies

Show which matters must resolve before other matters can resolve.

04

Find the next move

See the actionable matters whose blockers are already cleared.

Install it, then ask for your first next move.

Tell your AI agent: install matters.global from the official repository, then give it your first goal so it can build a map and suggest what to do next.

01

Install through your agent

Paste this exact instruction:

Install from the official repo:
https://github.com/matthiasroder/matters.global
02

Open your workspace

cd your-workspace
codex

# or open Claude Code in the same workspace
03

Use the system

Turn this into matters and tell me what is actionable now:
Goal: publish a first public walkthrough this week.
What should be done first?
Keep private context private.

Ask for the next move

Look at my matters graph and tell me what is actionable now.

Example next action

What is the single highest-impact thing I can ship next?

Start messy. Tell the agent what you are trying to understand or move forward. It will propose matters, conditions, and dependencies, then ask before adding them to your map. Your private context stays local; share only the public view you export.

A Sonophilia Foundation project for creative execution.

matters.global is part of the Sonophilia Foundation, a non-profit advancing and sharing creativity research through collaboration between science, business, education, and technology.

Creativity is the human capacity to solve unsolved matters in a new and useful way. We use that capacity to keep the focus on what is truly solvable, what is unknown, and what should happen next.

It was created as a Sonophilia Foundation initiative by co-founders Seda and Matthias Röder to build an interconnected map of all known problems. The project was started in 2020 and went through several phases with individuals from the Sonophilia Foundation, Ocean Protocol, and the Foresight Institute.

Born from Sonophilia's creativity and collaboration work

Shaped by cross-disciplinary task forces and community co-creation

Now open-source software linking goals, blockers, and AI-assisted next actions