Everything the business asked for, sized, valued, staged to go.
The sled is the queue: every piece of work anyone in the company asked for, one row each, with a size, a worth, and the seat it is waiting on. Work comes in from intake, gets sized and valued here, sits on the sled until it is dispatched, and comes back with a receipt. Priced before, receipt after, and the record underneath it is yours.
| stage | ask | size | band | worth | who | filed by |
|---|---|---|---|---|---|---|
| ready | SSO redirect loop on the Acme tenant | M | $4–11 · 20–55 min · 81% | high · Acme renewal | — | Samantha |
| ready | Rate-limit headers on the public API | M | $5–12 · 25–60 min · 77% | — (ask John) | — | Samantha |
| on the sled | Billing page: proration shown before confirm | L | $14–38 · 1–3 h · 64% | high · churn | Dana | John |
| on the sled | Export CSV respects the saved filter | S | $1–3 · 6–14 min · 90% | medium | overnight | Jake |
| on the sled | Empty state for a new workspace | S | $1–2 · 5–10 min · 93% | low | overnight | Samantha |
| landed | Webhook retries with backoff | M | $6.10 · 38 min actual | medium | Jake | John |
Size it before it runs. A letter to glance at, a band behind it.
/sled estimate in Claude Code, "size this" in a conversation, or nothing at all: a spec saved to the repo and a ticket opened in the tracker are sized on their own. Story points died when the volume rose and t-shirt sizes are the same guess in fewer letters, so the size here is derived, never typed — a letter for the glance, and behind it a cost band, a time band, which model fits, a confidence and a spend cap, calibrated from your own actuals per task type, per model, per repo.
› /sled estimate docs/specs/sso-redirect.md Size M · $4–11 · 20–55 min · 81% Fits Fable (band) · Sonnet (+40% time, ⅕ cost) · Haiku does not fit Worth high · $5k–20k · "Acme renewal" · stated in spec Cap $20 · nothing bills past it Suggest now · your seat has 38% of the weekly bar · resets Thu 9:00 Recorded as est-8c2f. The actual closes it on merge.
M is the glance; $4–11 · 20–55 min is the truth. The letter comes from the band, so nobody argues about points in a planning meeting again.
Which model fits comes from first-try CI pass, rework and parked-on-question rates for this task type in your record.
A spend cap rides on every estimate. An overnight job stops there; a session at the wall is told what is left.
Every actual moves the next band. The share of estimates landing in band is on the roll-up, for the workspace, never per person.
Set at intake by the person who asked. Adjusted by the lead. Learned over time.
Worth is value and priority on one gauge, and it is set in a fixed order. At intake, by whoever asked for the work, because they are the one who knows what it is for: a customer, a revenue line, a risk, or just high, medium, low. If they said nothing, the tracker's priority stands in at low confidence. On the sled, a lead can adjust it. Over time our model suggests it from that org's own stated and landed rows, always overridable. Tokens never become worth, and a missing statement is never a low.
| source | how it gets there | confidence | example |
|---|---|---|---|
| stated | A Worth: line in the spec, the ticket, or the PR body. In a conversation, one clause. Marked with who said it. | high | Worth: high · $5k–20k · Acme renewal |
| priority | The tracker's priority field, mapped to a level. | 0.25 | P1 → high · from Linear |
| suggested | A per-org model that learns from that org's stated and landed rows. Always overridable at intake. | low, rising | medium · 0.4 · suggested |
| measured | An outcome signal after the fact: an incident count drops, usage moves, a renewal closes. | as observed | incidents −60% · 30 d |
| unstated | Nothing said, nothing inferred. | — | blank stays blank |
The receipt: what came back, who built it, signed.
When the PR merges or the overnight job lands, the row gets its receipt: minutes, model or person, cost, and a signature over the job, the outputs and the time with the workspace key. Anyone with the public key checks it locally. Which model built something is a fact, not a claim, in a compliance review or a retro. The estimate it came from closes, and the next size is a little better.
Work done on a seat carries minutes, model and seat credits; work done on the platform key carries dollars. The three rails stay named apart in the record too.
SLEDS-Source: ENG-482 · spec §3 · linear Worth: high · $5k–20k · Acme renewal SLEDS-Estimate: $4–11 · 20–55 min · Fable fits · 81% · cap $20 SLEDS-Seat: now · Fable · jake@ · seat quota SLEDS-Actual: 34 min · $7 · fable-5.1 SLEDS-Attest: ed25519 · 2026-09-11T14:02Z · verified → ENG-482 moved to Done · est → actual written to the ticket
Day one: written back to your tracker. Day thirty: the sled is the record.
On day one sleds writes sizes, worth, actuals and receipts into Jira, Linear and GitHub as fields they already have, and replaces nothing. By day thirty the sled has seen every piece of work start, priced it, and signed what came back — which is more than the tracker ever saw — so the sled is already the record of the work and the tracker is a view over it that a growing share of the team stops opening. We keep it in sync for as long as anyone wants it, and it keeps every row if you leave.
| where | reads | writes |
|---|---|---|
| Linear | Issues on open; priority as a stand-in for worth; the intake label. | Estimate and worth on open; actual and receipt on merge, in the fields the tracker already shows. dispatch and intake file here. |
| Jira | Work items; "assign to an agent" as an input to dispatch. | The band and the actual, in Jira's own fields. Next. |
| GitHub | Every PR on open; the SLEDS-* lines in every shape they have ever taken. | The receipt on merge, including on our own PRs. |
| Slack | — | A digest of what landed, what is waiting, what hit a wall. Aggregates, never a per-person list. |