The Work Claim
When capability becomes personal, proof of work cannot mean handing over the whole private stack.
Stack People need legible contribution.October 9, 2026
01 / the missing proof
Not a screen recording.
A work claim is a short, reviewable account of what a person set out to do, what role their tools played, what they decided, and where another person can inspect or challenge the result.
For a long time, organizations used visible application activity as a rough proxy for contribution. Did the employee complete the form, update the shared system, attend the meeting, move the ticket, and leave a trace in the prescribed place? Those traces are imperfect, but they are easy to administer because the application owns the route.
The app-to-stack transition makes that shortcut weaker. A Stack Person may prepare a negotiation in a private workspace, compare options with a personal model, draft a plan through several iterations, and bring only a concise recommendation into the formal system. The important work may have happened before the approved screen ever opened.
The work claim is not a demand for a person’s prompts, private notes, or entire AI history. It is a bounded invitation to review the relevant judgment behind a consequential contribution.
That distinction matters. Managers and institutions have a legitimate need to understand responsibility, protect confidential material, and test important claims. People also have a legitimate interest in retaining the private contexts, methods, false starts, and working habits that make a personal capability useful. A good work claim gives each side a place to stand.
02 / the four lines
Make the relevant work legible.
A useful claim is smaller than a diary and more honest than a polished answer. It lets a reviewer see how a result should be trusted without mistaking total visibility for accountability.
- 01 / INTENT
Name the decision or outcome.
State the job the work was meant to advance: choose a vendor, frame a customer response, evaluate a risk, or prepare a recommendation. A review cannot begin if the purpose remains hidden inside a tool chain.
What was this for? - 02 / CONTRIBUTION
Say what the person supplied.
Distinguish the person’s framing, choices, and accountability from the assistance used for research, drafting, organization, or simulation. The point is not to pretend that tools were absent; it is to identify human agency clearly.
What did you decide? - 03 / BOUNDARY
Mark what did not travel.
Describe the relevant limits: no confidential material entered a personal system, a source remained internal, a decision required human approval, or a sensitive context was deliberately excluded. Boundaries are evidence of care, not an admission of failure.
What stayed protected? - 04 / REVIEW
Point to a checkable trail.
Offer the source, record, test, stakeholder, or decision point that can confirm or correct the claim. A reviewer does not need every intermediate keystroke; they need a real route to challenge the part that matters.
Where can this be tested?
03 / management after apps
Review the work, not the hidden life around it.
The managerial temptation will be to replace old application monitoring with an even more exhaustive demand for AI traces. That would confuse evidence with extraction.
Some work will need detailed records. Regulated decisions, safety-critical procedures, public claims, protected data, and formal approvals properly call for stronger documentation. A work claim does not dissolve those duties. It helps identify which record is relevant, who made the decision, and where review belongs.
But not every useful act should become a mandatory replay. If a person uses a private stack to understand a market, rehearse a difficult conversation, or find the first shape of an idea, forcing complete disclosure can make the capability less candid and less valuable. App People were often required to reveal themselves through the system’s fields. Stack People need a way to be accountable without becoming a permanently open interface.
This is one practical extension of The First Stack Generation: personal capability becomes economically meaningful only if people can carry a method across contexts while still making consequential work answerable to the people affected by it.
04 / a better bargain
Trust can have edges.
The point of a stack is not to make every worker opaque. It is to let capability grow around a person without requiring that all of the person be absorbed by the institution.
That requires a better bargain than either extreme. “Show us nothing” cannot carry a consequential decision. “Show us everything” turns responsible preparation into surveillance and makes private capability impossible to develop. The work claim occupies the productive middle: enough information to assess the contribution, not so much that ordinary thinking becomes corporate exhaust.
For Stack People, this will become part of professional craft. They will learn to state an aim before a system begins to act, to separate a recommendation from an automated draft, to keep boundaries explicit, and to leave a trail that someone else can follow when it matters. Those are not bureaucratic chores. They are the social technology that makes personal capability easier to trust.
For organizations, the question is equally sharp: are we asking for proof because we need to protect a real relationship, or because we only know how to manage work when every movement occurs inside our software? The first question builds confidence. The second recreates the app era with more data and less dignity.