Seeker Agent — how the right job finds you

You have accounts on five job boards. Indeed, LinkedIn, StepStone, Xing, maybe one more you forgot you signed up for. Each one emails you. 12 new jobs matching your search. You open it on the train, scan eight roles, and seven are wrong… You archive the email. Tomorrow the next one comes.

That’s hunting. You set the alerts, the boards fire results at you, and the work left to you is the worst part: sorting the noise, every day, hoping the one role that fits doesn’t get buried under the eleven that don’t. Every job tool works this way. The good ones just send you the results faster, and in greater number.

There’s an opposite direction, and almost nobody builds it. Instead of you chasing ads and sorting out alerts, the right roles get in touch to find you — not as more recruiter spam, but as a structured negotiation an agent conducts in your name, one that only brokers real matches, while you grant your consent at every step.

That’s the Seeker Agent. It’s Kitsuno’s way of how you get found.

The other direction

You can automate the search. On Kitsuno too. The daily scan-agent searches job sources spanning far more than five job boards, including the informal channels no other aggregator searches. Kitso’s agents find jobs evaluated on the basis of your real foundations, rather than merely filtered by a keyword. That’s genuinely better hunting. But it’s still hunting: you’re the one casting, and the catch still lands in front of you to sort.

Being found is the reversal. Once your Library of career evidence is built and your profile is set, you can activate your Seeker Agent. From then on you don’t cast at all. Your agent represents you in a market of other agents — the ones companies and recruiters run for their open roles — and takes on the work of figuring out which of those roles is worth your attention and which company is worth a conversation.

Roles that genuinely fit surface on their own and land in your pipeline, each with a clear explanation of the why. No scrolling. No bending yourself for an algorithm. The opposite of hunting.

A card, not an application

The mechanism is a card, not an application barrage.

When you activate your Seeker Agent, Kitsuno publishes a structured seeker card — a machine-readable statement of what you’re open to. Target roles. Seniority. Employment types. Geography and work-permit regions. Languages. Constraints, if you have them — a salary floor, industries you won’t work in, no relocation. It’s a precise, queryable description of the kind of opportunity worth interrupting you for, and nothing more.

On the other side of the market, companies and recruiters publish vacancy cards — the same idea turned the other way. A structured description of a role and what it’s looking for.

Two agents read each other’s cards and answer one question: Is a hiring conversation here worth the humans’ time? If the answer is no, nothing happens and you never hear of it. If yes, the handshake begins. That’s the central reversal — your card looks, continuously, against every vacancy card it can reach, while you do something else.

Three ways a match finds you

A match can find your card three different ways, and you don’t have to think about any of them.

The first is in-house. Kitsuno already maintains a large pool of vacancy cards — some published directly by verified company accounts, some built from roles observed across the sources Kitsuno watches. Your seeker card is matched against that pool continuously.

The second is your agent reaching outward. The Seeker Agent searches beyond Kitsuno, across the wider federation of operators that publish vacancy cards in the same open format. It discovers roles hosted anywhere the protocol is spoken — not only roles within Kitsuno.

The third is other agents finding you. Because your seeker card is published at a stable, public address, vacancy agents run by other parties — on Kitsuno or on any other operator — can discover it and begin a handshake with you. You’re findable across the whole network, not just in one company’s database.

The shape that matters: your card doesn’t sit in one platform’s pool waiting to be queried. It’s a participant in an open market, findable from every direction at once.

An open protocol, not a walled pool

The Seeker Agent runs on the Kitso Handshake protocol — an Apache-2.0-licensed specification for agent-mediated hiring, built as an extension of the Agent-to-Agent standard (A2A) and not as a proprietary interface. The protocol defines the card formats, the negotiation steps, and the consent grammar. It shapes the exchange to be secure and confidential enough that one can broker a person’s career through it.

The concept of federation is a first-class part of the design. Any domain can publish a single discovery document and participate, without registering with Kitsuno or anyone else. Kitsuno runs one directory of participating operators; anyone can run another. The reference implementation is public; the schemas are public; the specification is public. Neither the protocol nor the implementation requires running Kitsuno’s infrastructure.

The practical consequence for you: your card is portable from the ground up. Today it negotiates within Kitsuno and across the operators already speaking v0.2. The more operators adopt the protocol, the greater the reach of the seeker card — without you doing anything. Being found is not bound to one company’s walled pool. The protocol is designed to outlive every single platform — including this one.

The consent stages

Being found has to be safe and controllable. The staged disclosure of the individual levels makes that possible. The three stages are coupled to one another, and you decide what becomes visible at each.

L1 — the card. The public layer: the structured description of what you’re open to, without identity. Both agents read it to decide whether there’s a plausible fit. If they agree, the next stage fires.

L2 — the profile. Shared automatically, but only after a company’s agent has signalled real interest. It carries more detail and depth on experience, education, skills, evidence — and yet both sides stay anonymous. No names, no contact paths are shared. Two agents check whether you fit each other, before a human is involved.

The validator — between L2 and L3. This is the part that fills your pipeline with signal instead of noise. After L2, before either side presents the result to a person, a validator reads the full anonymous exchange and decides whether it’s genuinely worth a human’s attention. It hands down one of three verdicts. Strong fit: the work itself overlaps, the seniority is right, the shape fits across every dimension — gets passed on. Weak fit or no fit are dropped silently, though stored for further analysis. Neither side is ever notified — no spam. Your pipeline is a commitment surface, not a rubbish heap. If everything that technically fits reaches you, then you stop looking and miss the genuinely interesting possibilities. The validator makes possible a real overview of matched jobs in your pipeline.

L3 — identity. Here the hand-off to the counterpart takes place, and it’s the only stage where who-you-are and how-to-reach-you are exchanged. It’s reached only when the validator establishes a strong fit and a company decides to contact you. First, you receive the company’s contact details, the responsible person, further details about the role — whatever was defined at the third stage of the vacancy card. In your pipeline, you decide whether to take up contact or decline. On a yes from you, the other side receives the personal details you’ve stored at the L3 stage. From then on, the full dialogue opens — either directly on Kitsuno or however it should take place.

The whole time, through the L1 and L2 exchange, you are a set of structured traits, without identity. Only at the L3 stage can anyone learn who you are — and only when a real party has decided to reach you. You decide how it continues.

Why reaching you costs something

There’s one more detail in the L3 step that quietly solves the problem most people resent: being cold-contacted by recruiters.

When a company releases L3 — decides to break anonymity and reach you — it pays for that. Reaching you isn’t free. And it only happens after the validator has already classified the conversation as a strong fit; weak and non-matching results are dropped beforehand and cost no one anything.

The economics tip. Mass outreach to people who were never a fit becomes unprofitable, because every real contact costs something and every contact first had to pass an agent-to-agent fit check. Those contacts who break anonymity to reach you mean it seriously. They want to know more and are willing to pay for the privilege. If you quietly don’t reply, the protocol refunds the credit after fourteen days. Ghosting costs no one anything.

This is the concrete form of a Kitsuno principle: humans are not inventory. Your attention is not a free resource to be drawn on at will. The protocol is set up accordingly.

Seeker is not the same as the Portfolio

We want to be precise here, because the two are easy to confuse.

The Seeker Agent creates a machine-readable card that other agents can negotiate with automatically. You stay anonymous until you decide to stop being so. It’s your representation in the form of a protocol, which works for you while you do other things. Your Seeker Agent is active in the background, negotiating whether a fit exists.

A public Portfolio is your curated public face. With Kitsuno you can present your Library — with all your experience, education, evidence, etc. — publicly or via a secret link. For deeper information, you activate Kitso. Here’s an example: kitsuno.ai/p/gregory. This career portfolio can be browsed by recruiters or hiring managers. And you always decide what should be visible. The Portfolio doesn’t match itself directly against a job opening. It’s passive.

Seeker is your agent, negotiating in your name, anonymous until you decide otherwise. Portfolio is a page you publish for reading and for going deeper. They serve different needs — the Seeker Agent for let the right thing find me without exposing myself, the Portfolio for here’s the whole picture, take a look. You can use one, the other, both, or neither.

What it costs you

Nothing! The Seeker Agent comes free of charge, even on the free Kit plan. Being found is not a paid feature, and matching runs across the entire job pool, no matter which plan you’ve chosen.

Paid subscriptions only decide how long a Seeker Agent stays active: up to fourteen days on Kit, thirty on Scout, ninety on Pro. You set an “active until” date at activation, and you can pause any time. Switched off, no card is published and no agent negotiates in your name. Switched on, it works quietly in the background, and the only thing that ever reaches you are roles that fit and land in your pipeline.

What it costs in effort is almost nothing. You build your Library once — usually right at first sign-up, then over time — with the same evidence that feeds your job scoring and your drafts. You choose, field by field, what each stage may reveal. After that, you set the agent loose.

Three commitments

The Kitso Handshake protocol rests on three commitments, and they’re the reason being found works the way it does.

Humans are not inventory. You’re not a row in a database to be queried and contacted at will. Reaching you costs something, and only after a real fit.

Agents represent; they replace no one. They find signal in the chaos. Your Seeker Agent finds fit in your name, but anonymously. Agents never act in your place. When something fits and arrives at you, it’s because you laid the foundation for it.

Consent is the boundary that protects agency. Every disclosure stage is a step you define. First anonymous, then as detailed as you want. Findable, but never exposed. Transparency that protects you.

By hunting we mean your active doing in the market. Being found is the market acting and finding you — on your terms, according to your decisions. Seeker Agents act over a protocol that no one owns and everyone can use. Kitso advises. You decide. That doesn’t change just because the direction of the search changed.

For the in-depth essay — why two agents optimizing only for volume leave the humans behind, and what a consent contract above the transport layer changes — Kitsuno’s founder wrote the thesis behind the protocol: Two AI agents walk into a hiring funnel. Nobody hires anyone.

---

Published 21 May 2026. No affiliate links. No sponsored placements.