Observation
The problem
People sign up for our community sessions. Few come back.
Brought by the groupIllustrative
Bring the problem that keeps coming back.
Find a way forward, together.
Reason Commons is an open-source AI skill for working through what matters to you: understanding what is getting in the way, choosing a useful next move, and learning from what happens.
Bring your experience. Invite another perspective. Keep the reasoning in view as you decide what to try.
EarlyAvailable now as an early skill for Claude Code, Codex and compatible AI agents. Start on your own or with a small group.
You probably already have pieces of the answer.
- Someone has noticed a pattern.
- Someone else has a concern.
- You share a goal, but keep circling the same disagreement.
- Or you have tried a change and are still unsure what helped.
Reason Commons gives those pieces a shared place to work: a picture of what you are trying to achieve, what might be holding it back, and what deserves attention next.
A conversation you can act on. A record you can return to.
-
Bring something real.
A recurring frustration. A decision that has stalled. A conflict. Something you want to make possible. Begin with what you know.
-
Make the reasoning visible.
The AI helps connect observations, possible explanations and proposed changes. Contributions keep their names. Assumptions and objections remain available to examine. Nothing counts as agreed until people explicitly accept it.
-
Choose a next move.
Work towards a useful question or a change worth trying. Before a trial, record what you hope it will improve, what concerns you, and who could report on the experience.
-
Return with what happened.
Compare that experience with the original expectations. Keep what remains uncertain visible. Use it to work out where to go next.
What might that look like?
Illustrative session. These people and events are fictional.
Understand
Contributions keep their names. Assumptions and objections remain available to examine.
might explain: Maya's possible explanation, which follows, might explain the problem
Possible explanation
A possible explanation
“Maybe people leave without knowing how to get involved.”
MayaIllustrative
Concern
A concern worth keeping
“A more explicit invitation might feel pressuring.”
Attached to: a more explicit invitation, not yet proposed
RowanIllustrative
Try
Before a trial, record what you hope it will improve, what concerns you, and who could report on the experience.
Proposal
A small trial
End one session with a clear, optional next step. Ask newcomers how the invitation felt and whether they took it up.
- Proposed (current)
- Chosen (not yet)
- Tried (not yet)
Nothing counts as agreed until people explicitly accept it.
Not yet chosenIllustrative
-
What we hope it will improve
Possible explanation
“Maybe people leave without knowing how to get involved.”
MayaIllustrative
-
What concerns us
Concern
“A more explicit invitation might feel pressuring.”
Attached to: this trial
RowanIllustrative
-
Who could report
Newcomers: how the invitation felt, and whether they took it up.
Return
Compare that experience with the original expectations. Keep what remains uncertain visible.
A small trial
- Proposed
- Chosen
- Tried once (current)
An experience to return with
-
Recorded before · hope
“Maybe people leave without knowing how to get involved.”
MayaIllustrative
reported
Report
One newcomer found the next step helpful.
From a newcomerIllustrative
-
Recorded before · concern
“A more explicit invitation might feel pressuring.”
RowanIllustrative
reported
Report
Another felt put on the spot.
From a newcomerIllustrative
-
Recorded before · the problem
People sign up for our community sessions. Few come back.
Brought by the groupIllustrative
not yet known
Report · still uncertain
We do not yet know whether either will return.
Kept visibleIllustrative
- Observation: solid
- Possible explanation: dashed
- Concern: dotted
- Proposal: double
- Report: ruled
- Open question: open dashed
Your judgement has a place here.
-
Keep your contribution attached. Ideas, observations and challenges retain their contributors, so people can clarify, develop and build on them.
-
Keep room for disagreement. A concern can stay connected to the particular assumption or proposed change it challenges.
-
Keep expectations honest. Original predictions remain on record when experience comes back. Agreement is a working position; it does not establish truth.
A renaissance of common sense.
The ability to think clearly together should be part of everyday life.
By common sense, we mean a practice that keeps getting better: making assumptions visible, examining cause and effect, and letting experience change our minds.
Reason Commons draws on the Logical Thinking Process within the Theory of Constraints. You can begin in ordinary language, with something you care about.
Our ambition is a Reason Commons: people cultivating the capacity to reason together, contributing what they know, questioning what is missing, and helping one another make progress.
This skill is the first working tool for that ambition. The wider commons is something we want to build with others.
Help shape it by using it for something that matters.
We are inviting facilitators, community organisers and small-team leads to try the beta with one real problem, decision or aspiration.
Use it to develop a shared picture and a clearer next step. Then tell us what helped, what got in the way, and what happened when you tried something.
Your experience can shape what we build next. No background in the Theory of Constraints is needed. Today's beta is easiest if you or someone in your group already uses Claude Code, Codex or a compatible agent.
Help make good reasoning a public good.
As AI becomes part of how people think and act, we want tools that strengthen people's ability to examine a claim, challenge a recommendation and decide together.
- Today
-
We have an early working skill. Now we want to make it easier to use and establish where it helps in practice.
- What support could enable
-
We welcome conversations with public-interest funders, AI researchers and organisations willing to host pilots. Support could help us:
- Make access easier: reduce setup and facilitation burdens, improve accessibility, and test the experience with more kinds of groups.
- Build evidence: compare use with groups' usual ways of working, examine decisions and follow-through, and document friction and failure as well as useful outcomes.
- Develop shared foundations: improve open tools and documentation that others can inspect, adapt and extend.
- The question
-
The question we want to investigate is practical: does keeping reasoning, decisions and real-world experience connected help people make more useful next moves, at a reasonable cost in effort?
- What would count as evidence
-
Evidence of progress, for us, would include justified next steps, appropriate changes actually attempted, experience brought back into subsequent decisions, and people choosing to return. Those are questions for evaluation, rather than results we are claiming today.
A few things you might want to know.
- What is an AI skill?
- A set of instructions and supporting tools that gives a compatible AI agent a particular way of working. Reason Commons is currently installed into an agent such as Claude Code or Codex.
- Can I use it on my own?
- Yes. One person is enough to begin. A group can also work with one person operating the skill and naming contributions as people make them.
- Will the AI decide what is true?
- People examine and endorse the reasoning. The AI can misunderstand or propose a mistaken connection. Claims remain open to challenge; a convincing diagram is not proof.
- Does “commons” mean publishing my problem?
- “Commons” describes the ambition for the tools and shared practice. Your session's storage and sharing depend on your AI environment and the destinations you configure. The open-source tools and the reasoning you contribute are separate things.
- What does it cost?
- The skill's source is available under the MIT licence. Your AI provider's normal access or usage charges may still apply.
Bring what you are working on.
Tell us a little about yourself and what you would like to explore.