Built-in AI · Cross-product

An AI agent
inside every
Kubrik product.

Kubrik Copilot lives inside Kubrik OS, Kubrik POS, and Kubrik Commerce. It diagnoses problems with evidence, edits with preview-and-confirm, and reasons across modules in one turn — powered by 400+ tools and Claude, over the open MCP protocol.

How it works Try Kubrik Cloud
Kubrik OS Kubrik POS Kubrik Commerce
400+
AI tools
across 8 modules
3
RBAC tiers
consultant · executor · autonomous
<1.5s
First-token latency
p95 under nominal load
100%
Audit-linked writes
every AI write → ChangeHistory
Three RBAC tiers

Read. Confirm. Or auto-run safe writes.

Each user gets a tier per module. Tool catalogs filter what the LLM can see before the model decides what to do — the guarantee lives in code, not in prompts.

Consultant

Read-only diagnostics

The agent answers questions with evidence — what changed, what was rejected, what's missing — but cannot mutate state.

  • Trace primitives (readiness, attributes, prices, allocations)
  • Cites recent mutations from ChangeHistoryEntity
  • Cannot call write tools — they're never advertised to the LLM
Executor

Preview-and-confirm writes

Writes become proposal cards. Nothing mutates until the user clicks Confirm. Proposals persist if the tab closes.

  • Exact args + dry-run preview rendered inline
  • Tools tagged SAFE / REVERSIBLE / DESTRUCTIVE
  • Idempotent confirm — double-click = no double-write
Autonomous

Safe writes auto-run

Only SAFE-tagged tools run without confirmation. Reversible and destructive ones always surface a confirm card.

  • Notes, drafts, tags can land in one turn
  • Auto-action banner in the UI for every silent write
  • Audited row in change history with originating message
What it does

Three core moves — the rest is composition.

The same engine answers a why-won't-this-publish question, edits a SKU on confirm, and chases an order discrepancy across OMS, WMS, and Accounting.

Diagnose with evidence

Trace primitives return rejected candidates, override sources, recent changes. The LLM reasons over real data, with citations.

  • "Why won't SKU-1234 publish to ecom-eu?"
  • "Why is this price not used for customer X?"
  • "What changed on this order in the last 24h?"

Edit by natural language

"Set material to cotton on SKU-1234" becomes a preview card. The args are shown. Nothing changes until you confirm.

  • Bulk updates: "tag these 14 SKUs as clearance"
  • Status flips: "mark these 6 returns as approved"
  • Drafts: "create a draft promotion for next Tuesday"

Orchestrate across modules

The orchestrator calls tools across OMS, WMS, ACC, CRM in one turn. Users don't need to know which module owns what.

  • "Why is Order-10 short by 5 units?"
  • "Show me everything pending the close of Period-9"
  • "Which active promos still need approval?"
Capability lives in tool catalogs

400+ tools. 8 modules. One MCP transport.

Each module ships its own MCP catalog. The LLM can only do what its tools can do — that's the security boundary, not the prompt.

Kubrik Copilot · Platform
Session · MCP transport · Tool dispatcher · Write proposals

Per-tenant scope · per-user RBAC tier · per-module catalog · 20s per-tool timeout · prompt-injection delimiters · audit linkage on every write.

CORE
Products / PMS
48 tools
WMS
Warehouse
48 tools
OMS
Orders
50 tools
CRM
Customers
46 tools
ACC
Accounting
54 tools
DMS
Documents
42 tools
PSA
Services
56 tools
HR
Human Resources
68 tools
Open by design

Same tools, your own MCP client.

The in-app web terminal is just one MCP client. Connect Claude Desktop, your own scripts, or anything that speaks Model Context Protocol — with the same auth and the same audit linkage.

MCP-native

No black box, no lock-in.

Every tool is reachable from outside the app. Customers can automate against the same catalog the in-app agent uses — from a CI job, a scheduled report, or another LLM.

  • · Same auth: tenant-scoped, RBAC-tier-aware
  • · Same writes-as-proposals contract
  • · Same audit linkage to ChangeHistoryEntity
# Connect Claude Desktop to your Kubrik tenant
{
  "mcpServers": {
    "kubrik": {
      "url": "https://api.kubrik.io/mcp/sse",
      "headers": {
        "Authorization": "Bearer $KUBRIK_TOKEN",
        "X-Tenant-ID": "$TENANT_ID"
      }
    }
  }
}

Or ask from the AI you already use.

Kubrik exposes over 400 governed tools over MCP, the open Model Context Protocol. Point your own AI client at your tenant and it can interrogate the business and act on it — under the same permissions, thresholds and audit trail as everyone else. No export, no copy-paste, no separate BI layer.

ChatGPT Claude Cursor Copilot Studio any MCP client
Your AI client mcp://kubrik-os · tenant-scoped · read + write
Margin at the Centro branch fell almost 4 points last month. What happened?
acc.marginByBranch(branch="Centro", period="2026-08")
→ gross margin 22.4% · prior period 26.3% · −3.9 pts
core.marginByCategory(branch="Centro", period="2026-08")
→ Electrical −9.1 pts — 61% of the gap
pur.supplierCostChanges(category="Electrical", since="2026-07-01")
→ supplier ELEKTRA · +18.4% on 34 SKUs · effective 2026-07-14
ppe.priceListHistory(skus=34, since="2026-07-14")
→ 0 of 34 repriced
pos.unitsSold(skus=34, branch="Centro", since="2026-07-14")
→ 1,284 units · ARS 4.7M revenue
ELEKTRA raised cost 18.4% on 34 electrical SKUs effective 14 July. None of the 34 were repriced. Centro has sold 1,284 of them since — ARS 4.7M of revenue at the old price, which is roughly ARS 812K of margin you did not collect. Electrical explains 61% of the branch’s 3.9-point drop; the rest is product mix. The same 34 SKUs are also exposed at Norte and Depósito Sur.
Reprice them to restore the previous margin.
Write proposal · awaiting your confirmation
ppe.applyCostIncrease(supplier="ELEKTRA", skus=34,
  target="restore_margin", effective="today")
Simulation first: +18.1% average list price · margin restored to 31.2% · 4 SKUs would cross a psychological price point and are held back for review.
Confirm Reject
Autonomy level 02 — prepare. Price changes sit above this role’s threshold, so the agent builds the transaction and stops. Raise the rung and it would apply within your band on its own; either way the write lands in the audit trail with the tool call, the inputs and the reasoning attached.

Illustrative example with representative figures. The tool names, the call sequence and the governance behaviour are the real ones.

Available in every Kubrik deployment.

Kubrik Copilot ships with Kubrik OS, Kubrik POS, and Kubrik Commerce. No separate plan to buy, no add-on to enable — just an embedded AI layer waiting for the first question.

Start a sandbox → View pricing