Linkubit
Control

Rules, knowledge and approvals for an AI agent

The agent answers from what you gave it, does only what you allowed, and asks before anything that costs money.

In short

Linkubit's assistant answers from the documents and catalogue you gave it, and is held to rules that run in code: a gate before any tool call, a gate before any outbound message, and a verifier on every reply. Anything that spends money — a campaign, a budget, a price change — waits for a person's approval.

Governance that is not a prompt

Rules enforced in code, not prompts

A prompt instruction is a suggestion. Rules live outside the model at three points: a gate before any tool runs, a gate before anything sends, and a verifier on every reply.

  • A rule the model cannot talk its way past
  • Every refusal is recorded in the audit log

Never quote before lane and volume are captured

Tool blocked

Hazardous cargo always goes to a human

Escalated

Discounts above 10% need a manager

Approval required
Enforced at three points outside the model: before a tool runs, before anything sends, and on every reply.

Money needs a human

Approvals for anything that spends

Requested ad campaigns, budget moves, discounts past a limit. A person decides, with the number attached.

  • The agent can request, never apply
  • Thresholds are yours to set, per rule
Approvals
1 waiting

12% discount on quote Q-1148 · ₹4.2L

AI

Requested by the agent mid-conversation. Above the 10% threshold, so it stops here.

ApproveDecline
A person decides, with the number attached — never applied silently.

The rest of control

Five effects, not free text

Block a tool until fields are captured, require a human first, force escalation on a topic, require approval above a threshold, or append an instruction. A closed set is auditable; free text is not.

The agent will not deny being an AI

It does not volunteer it either. Denial is refused by the verifier, outside the model, and is not configurable.

Grounded knowledge with citations

Chunks link back to the document they came from, so an answer can point at its source instead of asserting something unbacked.

Workflows as rows

Trigger, condition, action — edited like any other configuration. A visual canvas is a product in itself and is deliberately deferred.

Give me 20% off and I’ll book today.
Rule · enforced in code
blocked
Never offer a discount. Hand a discount request to a human.
Draft reply “I can do 15%…” was never sent
I can’t change the price myself — Priya from our team will message you in a few minutes about it.Agent
Guide · 6 min · Updated 19 September 2026Rules an AI agent cannot break: why a prompt is a suggestion, and how a gate before every tool call and every send makes a rule realA rule in a prompt holds most of the time. A rule in code — a gate before each tool call, a gate before each send, a verifier on each reply — holds every time. What that looks like, the rule effect that does nothing, and why the gate fails closed.
This is the third time the delivery has been late. I want a refund.
Handed to Priya · refund requestagent paused
Hi, this is Priya. I’m sorry — that’s on us. I’ve refunded ₹1,299 to your UPI and added a note for the next order.Priya · human
Guide · 5 min · Updated 19 September 2026When an AI should hand a WhatsApp conversation to a human: the triggers, the handover that works, and what not to hand overHand over on a rule — a topic, a value, a missing capability, sentiment, a flood — before the reply, not after a wrong one. What the person needs to see, why the agent stays silent until handed back, and the ten percent that should be a person's.
A keyword bot
Price for the batch starting in October?
Sorry, I didn’t understand. Reply 1 for courses, 2 for fees.Agent
An AI salesperson
Price for the batch starting in October?
The October CAT batch is ₹38,000, weekends. Are you a working professional or a final-year student?Agent
Guide · 6 min · Updated 19 September 2026AI sales agent vs WhatsApp chatbot: what a flow, an assistant and an agent each do when the conversation leaves the scriptA chatbot follows a flow and fails outside it. An assistant answers from documents. An agent pursues a goal — qualify, quote, book — with rules enforced outside the model. Which one a business needs, and the failure mode of each.

Ready to put your sales on autopilot?

Start free

No coding required. No credit card required.