Vlynks

A Vlynk is an intelligence module. It reads the cards on your board and writes you an analysis — and what it finds also reaches the AI agent.

What a Vlynk gives you

Twelve Vlynks ship today. Every one writes a card on your board when it runs — an analysis with tabs, charts where the data supports them, and the cards each finding came from. The AI agent also reads each Vlynk’s last run (and when it ran), so you can ask about it in chat.

They reason in one of two ways, and each Vlynk’s contract says which:

• Rule-based — Client Health Monitor, ROI & Break-Even, Customer Intelligence and Opportunity Detector. Fixed, documented arithmetic and rules over your cards. No AI call; the same board always gives the same result.
• AI-written — Risk Calculator, Financial Control, Strategy Commander, Institutional Memory, Content Gap Analyzer, Morning Brief, Channel Performance Monitor, and Operator Actions. The AI reads your board’s cards under the Vlynk’s published contract. The same rules apply to every line: it cites the cards it came from, a number appears only if a cited card states it, and a line that claims a cause is removed. Each run is an AI request and counts toward your plan’s AI usage.

The contrast with a source stays exact:

	Source	Vlynk
Produces	A live card fed by an external account	An analysis card read from the cards on your board
You interact with it	Directly — you see the card	Run it, read its card, or ask the agent about it
Answers	“What is my data?”	“What does my data mean?”

A source tells you revenue was £4,200 last month. A Vlynk tells you that is the third consecutive month of decline — and shows you which cards say so.

What ships installed

Institutional Memory only. It is the one Vlynk every account starts with.

The other eleven are opt-in from the Marketplace. This changed in 0.4.27 — earlier versions pre-installed several, and we stopped, because intelligence you did not ask for watching data you did not choose is not a good default.

Installing

1. Open Marketplace in the rail
2. Find the Vlynk
3. Install

It starts working immediately. Most Vlynks depend on particular sources, and a Vlynk with nothing to watch stays quiet rather than inventing something — check its dependency before wondering why it is silent.

The Free plan holds three

Plan	Vlynks installed at once
Free	3
Plus	Unlimited
Pro	Unlimited

Reaching the limit shows an upgrade prompt. Nothing is uninstalled on your behalf — the limit blocks the next install, and anything already installed keeps running.

See exactly what a Vlynk does — before you install it

A Vlynk reads your data and shapes what the agent tells you. That is a lot of trust to hand over on the strength of a name and a one-line description, so every Vlynk publishes an Intelligence contract you can read.

Open it with View contract → on the Vlynk, in the Marketplace or the Vlynk panel.

What the contract shows

What it does — a plain description of the job.

Reasoning prompt — the exact reasoning this Vlynk applies to your data. Not a summary of it, not a paraphrase: the actual instruction it follows. If you want to know why a Vlynk reached a conclusion, this is where the answer is.

Data access — split into two lists:

• Reads — what it can see
• Cannot access — what it is not able to reach

The second list is the one worth reading. A Vlynk that watches your revenue telling you plainly that it cannot see your calendar is a stronger statement than any assurance about it.

Published by — who wrote it.

It is view-only, and that is the point

The contract is read-only. You cannot edit the prompt, and neither can anything else at runtime — it is published with the Vlynk and shown for transparency.

That is what makes it worth reading: a prompt you could change would tell you nothing about what the Vlynk actually does, and a prompt that could change itself would be worth even less.

Before you install

Open the contract first. Two questions it answers quickly:

1. Does it need data I have not connected? A Vlynk whose Reads list is full of sources you do not have will stay quiet rather than do anything useful.
2. Am I comfortable with what it sees? The Reads list is the honest answer, not the description.

The twelve

Institutional Memory (installed by default) — the board’s historian: the decisions made, the threads left open, and the context you need today, from the board’s change history and its cards.

Operator Actions (Plus and Pro) — a daily operating record of your whole workspace and five prioritized moves, with an outcome loop. It has its own page.

Morning Brief — reads every card on the board to surface what matters today: the top signal, the one action worth doing, and what to watch.

Opportunity Detector — revenue, lead and retention openings from the cards on your board — connected or manual — with confidence, the expected outcome and one action each.

Client Health Monitor — merges messages, bookings, deals, payments and reviews into per-client health: who needs attention, and why.

Customer Intelligence — customer health, conversion funnel, revenue concentration and retention trend from the client data on your board, including client spreadsheets.

ROI & Break-Even — break-even point, ROI, cost breakdown and runway from the revenue and costs on your board. Costs are read wherever you write them — a note, a list, a text card or a spreadsheet (“rent $2,000, salaries $8,000”).

Financial Control — budget against spend, where the money goes, and your cash position — every figure taken from a card.

Risk Calculator — scores the risks building up on the board — deadlines, blocked work, money owed, clients going quiet, scope — each with its evidence cards and one way to reduce it.

Strategy Commander — reads the board as a strategy map: your priorities, how well the work lines up with your goals, and the next moves.

Content Gap Analyzer — content coverage gaps between your channel and the competitors you track, using public video and channel signals.

Channel Performance Monitor — “how is my channel doing?”: views, subscribers and their velocity, and the growth levers your own numbers support, from your YouTube channel and analytics cards.

Running them

• From its card — Refresh runs it again on your board as it is now.
• From the agent — ask for it by name: “run the Risk Calculator”.
• On a schedule — set the card to run Daily or Weekly. AI-written Vlynks start with the schedule Off, because each run uses your AI allowance.
• On their own — Operator Actions writes once a day, the first time you open Vexlynk; Morning Brief runs the first time you come back after six or more hours away.

A Vlynk with nothing to read says so, and names what would help, rather than inventing a finding.

Next

• Using the AI Agent — how Vlynk intelligence reaches you
• Connecting Sources — give your Vlynks something to watch