Skip to content

Reporting Requirement

Reporting protocol — instantiates Principal–Agent Alignment

Obliges the agent to surface status, exceptions, and bad news on a fixed cadence, turning the agent's own account into the principal's first window on the work.

A Reporting Requirement aligns an agent by obliging them to keep the principal informed — a standing duty to surface status, progress, exceptions, risks, and bad news on a defined cadence and in a defined form. Its defining trait is that observability here is push, not pull, and self-authored: the agent volunteers the account rather than the principal going to look. That makes it cheap and timely — the principal learns of trouble as the agent sees it, early enough to intervene before drift becomes failure — but it also means the window is the agent's own, and its honesty is the mechanism's central vulnerability. A good reporting requirement is engineered to produce usable evidence and early warning, not paperwork: it asks for the exceptions and the bad news specifically, because those are exactly what an agent left to their own devices would rather not mention.

Example

A private-equity firm owns a controlling stake in a portfolio company it doesn't run day-to-day. The firm needs to know how the company is tracking without stationing someone on-site, so it imposes a monthly reporting requirement on the CEO. By the tenth of each month the CEO must submit a standard pack: financials against plan, a short narrative on what changed, a risk-and-exceptions section calling out anything off-track, and any covenant or cash-flow concerns flagged explicitly rather than buried.

The requirement's value shows in a soft quarter. Revenue is running behind, and the reporting format forces the CEO to surface it in month two — with the reasons and the plan — rather than letting the firm discover it in month four when options have narrowed. The report doesn't verify anything (the CEO writes it), and that's its limit: it is the firm's first and fastest window, giving enough signal to ask sharper questions, tighten the next check-in, or trigger a deeper look. It buys early warning, not proof.

How it works

The requirement is specified as cadence, content, and format. Cadence sets the rhythm (monthly, weekly, on-event); content names what must be reported — and the design work is in demanding the uncomfortable parts (exceptions, risks, conflicts, uncertainty, bad news) by name, since a report that only asks "what went well" reliably gets it; format standardizes the pack so it's comparable across periods and hard to bury a problem inside. What makes it this mechanism rather than an audit is the direction of flow: the agent is the author and the source. The best versions lower the cost of honesty — a blame-free exceptions section, a norm that early bad news is rewarded, not punished — because the whole thing runs on the agent's willingness to tell.

Tuning parameters

  • Cadence — how often reports are due. Frequent reporting gives fine-grained early warning but burdens the agent and can devolve into box-ticking; infrequent reporting is light but lets problems mature unseen.
  • Content demand — how much the format forces disclosure of exceptions and risk versus leaving it to the agent's discretion. Explicit demands surface bad news; open-ended ones let it go unmentioned.
  • Structure rigidity — fixed template versus free narrative. A rigid template makes reports comparable and gap-visible but invites rote filling; a loose one captures nuance but hides omissions.
  • Verification coupling — whether reports are ever checked against independent evidence. Loose coupling is cheap and trusting but game-able; tight coupling (spot audits of reported figures) deters shading at the cost of effort and signaled distrust.

When it helps, and when it misleads

Its strength is timeliness at low cost: the principal hears of trouble early, from the person closest to it, without paying to independently watch. It makes the relationship resilient — drift becomes visible while it is still correctable — and it forces the agent to organize and confront their own status.

It misleads because the agent controls the narrative. Self-reports are shaded optimistically, bad news is delayed to the last defensible moment, and a slick report can substitute for substance — the agent who is good at reporting looks aligned whether or not they are. It also decays into ritual: a monthly pack nobody reads and nobody checks trains the agent that content doesn't matter. It only narrows the information asymmetry between principal and agent; it does not close it, because the informed party is still the one holding the pen.[1] The guard is to demand the exceptions explicitly, make honest early warning safe to give, and couple the reports to occasional independent verification so the account can't drift freely from the facts.

How it implements the components

  • information_reporting_channel — it is the recurring channel: the defined cadence and format by which the agent surfaces status, exceptions, conflicts, and bad news to the principal.
  • observability_and_monitoring_plan — the agent's self-disclosure is one leg of the observability plan, the push-side complement to independent examination.

It does not independently verify what the reports claim — that is Audit or Review Cycle; it does not carry the urgent out-of-bounds interrupt — that is Escalation Protocol; and it neither defines the metric targets (Service-Level Agreement) nor imposes consequences on what it reveals (Governance Board).

  • Instantiates: Principal–Agent Alignment — implements the information-flow component that keeps the principal informed enough to intervene.
  • Sibling mechanisms: Audit or Review Cycle · Escalation Protocol · Governance Board · Service-Level Agreement · Performance Contract · Decision-Rights Matrix · Incentive Compensation Plan · Reputation System · Fiduciary Duty Rule · Clawback Clause

References

[1] Information asymmetry — the condition in which the agent knows more about their own action and situation than the principal does. A reporting requirement narrows the gap by obliging disclosure, but because the agent authors the report, it cannot fully close it; independent monitoring is needed for that. ↩