In developmentShared decisions for AI coding teams

Keep decisions across AI coding sessions.

For designers and developers using AI coding assistants, a choice made in one chat can get lost in the next. Airouz is exploring a shared log of human-approved decisions that agents can consult before they build.

Product concept in development. The workflow and interface below illustrate the intended product; they are not a live application or a claim of available integrations.

Planned clientsClaude CodeCursor
Planned workflow

From one chat to a shared decision.

The goal: let a person approve a decision once, then make it available to other agent sessions. The steps below describe the proposed workflow.

SOURCEAgent sessionClaude Code · Cursor · any MCP
DRAFTProposed decisionairouz.propose
HUMANOwner confirmsin session · Slack & emailSoon
REPOdecisions/Markdown, reviewed in PRs
EVERY SESSIONContext & checkbefore and after each task

Planned flow: an agent proposes a decision, a person approves it, and other sessions consult the saved Markdown before checking their work.

Product direction

A small, strict surface.

We are designing three MCP tools: propose, context and check. MCP is the connection through which an AI assistant could read approved decisions and return possible conflicts for human review.

Propose · planned

The agent drafts a decision with its reason and source. It stays a draft until its owner confirms it.

Context · planned

Before a task, a session loads only the confirmed decisions that touch what it’s about to change.

Check · planned

After a task, output is compared with confirmed decisions. A contradiction comes back with the decision and who owns it.

In your repo

A decision file you can review.

The proposed storage format is Markdown in your repository, so decisions could be reviewed and versioned alongside code. This fictional example illustrates the format under consideration.

decisions/D-014-sign-in.mdconfirmed
  1. ---
  2. id: D-014what agents cite
  3. status: confirmedonly a person sets this
  4. owner: designwho to ask
  5. confirmed: 2026-10-07T10:42
  6. source: claude-code/7f3athe session it came from
  7. expires: 2026-12-31retires instead of going stale
  8. ---
  9. Sign-in uses magic links. Do not add a password field.
  10. Why: fewer reset tickets; one less field on mobile.
Why Airouz

Designed around human approval.

The design goal is to preserve deliberate choices, their reasons and their owners, so a future session can distinguish an approved decision from a suggestion.

The problem to solve
Planned Airouz approach
Who can write
Suggestions mistaken for decisions
Agents propose. People confirm.
Where it lives
Context scattered across chats
Markdown in your repo
Old context
Old decisions lose context
Expires, with an owner to ask
Contradictions
Conflicts discovered late
Flagged when the task ends
Teammates without an agent
Approval trapped in an agent session
Confirm from Slack or emailSoon
Roadmap

Where we are.

Airouz is an early-stage, self-funded project started around June 2026. It is not incorporated. The landing page presents the product direction; public product access and integration validation are still ahead.

Current stage
  • Product concept and workflow design
  • Illustrative interface and decision format
  • Seeking feedback from prospective users
Planned first implementation
  • MCP tools: propose, context and check
  • Human-approved Markdown decisions
  • Validate with Claude Code and Cursor
Further exploration
  • Slack or email approval
  • Expiry reminders and shared log views
  • Checks across repositories
FAQ

Questions teams ask first.

Does Airouz see our code?

Checks are designed to run inside each person’s own agent session, and decisions live in your repo. Before anyone installs it, we’ll publish exactly what leaves your machine.

Which agents does it work with?

Claude Code and Cursor are the first intended clients. The planned Claude workflow is to propose a decision, ask a person to approve it, retrieve relevant Markdown decisions through MCP, and check task output for possible contradictions. Compatibility and reliability still need to be validated; this is not an announced live integration.

What if an agent proposes something wrong?

The intended rule is that an agent can propose a draft, but a person must approve it before other sessions treat it as a decision. This approval mechanism is part of the planned implementation.

Does everyone on the team need an agent?

We are exploring Slack or email approval for teammates who do not use a coding assistant. These integrations are planned, not currently available.

What happens to old decisions?

The proposed format includes an owner and an optional expiry date. Renewal, retirement and reminders remain design goals.

How much does it cost?

Pricing and public availability have not been announced. Contact us to discuss the problem and hear about future access.

Talk to us

Stop re-deciding the same things.

We’re looking for a few small teams where designers and developers build with agents together. Tell us where your decisions drift.

An independent, self-funded project. Started around June 2026; not incorporated.