For when the queue exceeds the bench.
Season overflow capacity for tax firms — and paid work for the bench when the season ends. One routing market, entered as a firm.
A tax firm cannot be the right size. The bench that survives April is too big for June; the bench a firm can afford in June is too small for April. Every year the same two failures arrive on schedule, and every managing partner knows both prices by heart.
The demand is not one season — only the staffing model is. Estimated-tax obligations fall quarterly by statute, extension filings run into autumn, amendment and notice clocks land all year. The cell’s sibling deck posts the same fact to the individual credential; this deck posts it to the firm’s P&L: the queue’s shape and the bench’s shape disagree by design, and that disagreement is a routing problem, not a hiring problem.
The villain is not the staff, not the clients, and not AI. It is the season itself as a staffing model: a year of practice economics compressed into one calendar quarter, priced against a fixed bench that can only ever be wrong in one of two directions.
The fin door grid files this name as two rows and one door: a firm buys capacity when its queue exceeds its bench, and a firm posts idle bench when it doesn’t. Those are not two products — they are the two sides of one routing market, and the same firm sits on both sides of it in the same year, just in different months. The bench is the unit on both sides.
And the candour a staffing agency never offers: firm-to-firm swaps alone cannot clear this market, because every tax firm peaks the same April. What decorrelates it is the cell this door belongs to — the individual pools its sibling doors recruit (the retired, the between-roles, the off-season credential holders of gigs.tax; the professional preparers of preparers.tax) and the estate’s year-round demand for tax acts. In season, the firm buys from pools that do not share its peak. Off season, the bench sells into demand that does not share its trough. That is why this is a door of a regulated cell and not a temp desk.
the firm is never a marketplace actor: individuals remain the actors, each verified under their own credential; a firm never claims a file — a named member does; no invisible substitution mid-file; firm standing is derived from members’ verified records, never entered by hand
machineryThe ruling is ported from the estate’s other firm door and holds harder here, because the acts are reserved: a return prepared for compensation carries a paid-preparer line that names a person, and sign-off judgment belongs to a named credential. The envelope carries the flag; a person carries the signature. That is the law’s architecture, and this door is built to match it rather than blur it.
The named seat is statutory, not sentimental: Section 6695 imposes penalties on a paid preparer who fails to sign the return or furnish an identifying number — the law requires the named person, then fines the absence. A market that routes firm overflow must land every such act on a named, verified credential; this one is designed to be unable to do otherwise.
Taxpayer data is criminally protected: Section 7216 sanctions a preparer’s knowing or reckless disclosure or use of return information outside the engagement. A file crossing a firm boundary is exactly where improvised subcontracting fails; here the consent and data-handling machinery is the cell’s burden by design — carried for the firm, never delegated to it.
The cell’s fee discipline is already posted on its individual doors: flat, fixed at post time, disclosed pre-claim, never contingent, never a share of a refund — the shape Circular 230’s contingent-fee restrictions already point at. This door inherits that discipline on both sides: a firm buying capacity sees the flat fee before any file routes, and a bench member working a routed file is paid the same flat fee whatever they find in it.
The discipline is rule-shaped, not a perk: Circular 230 restricts contingent fees for matters before the IRS, with only narrow examination-context exceptions. A capacity market whose fees never key to refunds or outcomes is the market the practice rules already point at — on the demand side and the supply side alike.
What capacity costs and what the bench earns are your first questions, and they deserve numbers, not adjectives. No fee figure of any kind appears in this record until the fin packet names the firm-bench SKU and real files price it — capacity fee band–·–posts when firm-bench SKU + first live fee data resolves — and the deck says so rather than inventing a rate card. The fin offer ledger files a season-overflow desk as entity-gated roadmap intent; that is a filing, not a price.
firm managing partner
both sides of the firm routing market — this record
CPA / Enrolled Agent
credentialed sign-off judgment — serves today
professional tax preparer
preparation work and the paid-preparer seat — filed, register leaf
Enrolled Agent
Circular 230 representation — filed, register leaf only
work-product name
amended-return work; queued to alias into preparers.tax unless a distinct SKU is named
A brand here is one ICP and one motion. The individual credential selling judgment walks through gigs.tax; the professional preparer walks through preparers.tax; the firm responsible for other people’s paychecks walks through this one. Same cell, same credential lattice, same fee discipline — different actor, different machinery. If you hold the credential and it’s your own engagements you want, one of those doors is yours, and this deck just told you so.
gigs.tax serves. The cell’s judgment door is live with its early-access funnel, and its record carries the same cell gates this one does.
preparers.tax serves the estate’s own RESERVED register leaf — “Reserved for the preparers’ door of the tax work,” with the words “Nothing at this domain is live.” The preparation door is filed, not launched, and nothing green here says otherwise.
The cell’s demand rail is not yet named on any door, and this deck does not pretend it is. It posts here — name, record, cross-link — when it is named and ratified, not before.
overflow.tax serves the estate’s own RESERVED register leaf — the live role sentence reads “Reserved for overflow season capacity for tax work,” and the register row reads “Season overflow routing to credentialed preparers when a firm’s queue exceeds its bench. Nothing at this domain is live.” The namespace position is occupied and honest; the door is filed, not launched.
The register the leaf files itself under is live: apis.finance serves the family hub today. The one-way seam is worn openly — the hub’s own published family map names apis.finance, apis.credit, apis.loans, and apis.mortgage, and does not yet name this door; the filing is attested leaf-side only, and reconciling it one way is queued in the program decision queue.
The live leaf’s sentence names only the demand side — capacity when the queue exceeds the bench. The fin door grid rules the two rows one market and queues widening the sentence to say so — a firm posts idle bench; a firm buys capacity. Until the leaf carries both sides, this deck quotes the narrower sentence that actually serves, and files the wider one as what this record means.
If nothing changes: the firm keeps being sized wrong twice a year — files turned away in March with the revenue attached, trained people idle in June with payroll attached, and the annual gamble on unverified seasonal hires signing under the firm’s name.
If it works: an April staffed by verified credentials at a flat fee known before any file moved, and a June in which the bench earns under the firm’s own flag — a firm sized to its year instead of its worst week.
The candour under everything above: the fin canon has not yet ratified the managing partner into its persona slate, and no firm-bench SKU exists in its offering stack. This record is filed ahead of both on purpose — the door grid’s re-opening condition, worn as a gate rather than implied away.
The fin canon rules that no .tax persona door ships root-surface copy until its voice spine exists — a spine clears copy, not a build. This record is filed ahead of that spine on purpose, and the root stays a register leaf rather than fabricate a funnel. The product surface posts when the spine clears it, not before.
The operating entity is designed, not formed — the same entity, and the same gates, both sibling doors post. No file routes across a firm boundary before the entity, its authorizations, and its coverage exist. Filed, not launched — verification is real before the word “verified” is.
enrolled firms–·–posts when stack#1 §A5 resolves · bench depth–·–posts when stack#1 §A5 resolves · routed file volume–·–posts when stack#1 §A5 resolves — no figures are presentable until the numbers gate resolves. No firm count, pool figure, or utilization number is asserted anywhere in this deck: the market is measured after the entity forms, never estimated before it.
This door is filed ahead of its funnel, and the honest ask matches: the root surface is the estate’s own register leaf, live today, with its contact line posted on the page itself — no signup theater before the cell’s gates clear. When the firm roster opens, firms enrolled before routing begins are designed to hold first-capacity priority in their states and return types, and the founding cohort shapes the capacity terms before they post rather than being announced to them. If your credential is your own and it’s your own engagements you want, walk one door over — gigs.tax serves today, and preparers.tax is the preparers’ door of the same cell.