Private pilot·The Governed Context Edge is live and onboarding design partners now — one governed canon your whole team's agents agree on. (Your personal layer stays free.)
For teams · the Governed Context Edge · private pilot
One governed canon for your whole team's agents. The Governed Context Edge connects your personal context layers: propose the definition you reconciled up to the shared canon, and adopt the team's canon back down with agency — adopt it, override it locally, or fork it, never a silent overwrite. It's built and live in a private pilot, onboarding design-partner teams now. And the solo layer you install today stays free — your own context, your own agent, no team account.
Connect your AIPrivate pilot
Design-partner terms: we onboard you personally, we set the price together, and you shape the roadmap. A handful of slots — that's deliberate.
The mechanic
The hard part of shared metrics was never storage — it's the relationship between one person's definition and the team's. ClariLayer's answer is the context edge: a governed connection between two context layers that says what flows, in which direction, and under whose authority.
The definition you reconciled in your own agent can be offered to the team's shared layer — with its SQL, its provenance, and the reasoning behind it attached.
When the team has a canonical definition, it flows back to your layer — but you decide: adopt it, override it locally, or fork it. Nothing overwrites your context silently.
Every edge records who can change what, which direction wins on conflict, and how it's audited.
Finance, marketing, and sales ops have legitimate reasons for different views of the same metric. The problem was never disagreement — it's invisible disagreement. ClariLayer keeps the shared core visible while recording which variant is current for which purpose. A visible, governed disagreement beats a fake consensus that breaks in the next board meeting.
Freshness
A definition doesn't stop being right the day the team agrees on it. A table gets renamed, a join changes, someone edits the source model. When you pull a shared definition into your agent, ClariLayer checks it against the team's canon and flags when your copy has drifted — you catch the divergence in flow, before a dashboard quietly serves the wrong number.
It leans on the same reconcile you use solo: in the personal MCP path, your agent runs the check with its own access and sends the supported evidence to ClariLayer. A mismatch surfaces as a caveat. The shared canon carries a freshness signal today — when a canon entry has gone stale, your agent sees it flagged the moment it recalls it. Scheduled re-verification cadence and owner paging are what the pilot is building with design partners. The separately gated team pilot can use an explicitly configured warehouse connector under its own access and operational controls.
What crosses the edge
A definition crossing an edge isn't a bare string. It carries its owner, its version, its validation evidence, and the reasoning behind it — so whoever adopts it (a teammate, or their AI data agent reading the governed contract) knows whether to trust it.
Not every shared definition needs the same rigor. ClariLayer matches the level of rigor to the stakes:
Tier 0
Open to anyone, self-serve release — for internal experiments and exploratory analysis.
Tier 1
Open to anyone, with warehouse validation before release and optional owner approval — for day-to-day team metrics.
Tier 2
Reserved for designated owners with multi-approver review — for board-level reporting.
Every iteration creates a new version instead of rewriting the old one — the evidence is never lost.
The conversation audit trail is institutional knowledge that cannot be recreated. After 12 months with 50+ governed metrics, migrating means losing the reasoning behind every definition.
“Why did we exclude refunds from MRR?” — the actual reasoning chain, not a changelog entry.
“Who approved this churn definition?” — the full approval chain with timestamps.
“When was this last validated?” — validation evidence attached to every version.
Private pilot · early access
Early access is hand-run. Tell us about your team and the definition that's already causing arguments, and we'll set up a pilot with you — no self-serve signup, no price on a page.
For teams · private pilot
The product you install today is the free single-player context layer — it connects to Claude Code, Cursor, or Codex over MCP so your agent stops re-explaining your data. The Governed Context Edge is the team strand: built, in private pilot, and hand-set-up with a few design-partner teams. If a shared definition is already causing arguments, request early access.
We use privacy-friendly analytics
With your consent we use PostHog and Vercel Analytics to understand how ClariLayer is used so we can improve it. We never sell your data. Errors are always monitored (without analytics) so we can keep the app reliable. You can change your mind anytime.