on the engineer's side

Your laptop. Your install. Your numbers.

sleds runs on your laptop, installed by you, as you, in one line. That is a privilege, and the product only keeps it if you are certain it works for you. Token spend is a number that could grade a person; we refuse to let it. This page is the charter every feature is held against.

Install What sleds never does

1line insleds uninstall is one line out. Hooks, status line, token, skill, Codex provider — every trace. Your tracker keeps the estimates and actuals.

0leaderboardsNo per-person spend lists, no CSV export that becomes one. Leads see aggregates by team and area.

yousee your numbers firstIn full, always. A manager sees them only when the workspace turns that on — and you are told.

what it does for youabout the work, not the money

The best way through your list.

Most engineers never touch credits and rarely see a wall. What every engineer has is a list and a limited day. sleds's job is the best way through it: which items are ready and which are blocked; what to run now and what the agent can do in the background while you think; which model fits which step — plan on the fast one, build on the strong one, review on the cheap one, by fit, not price; what is worth running overnight so it is done in the morning; what should go to a teammate who owns that area. Your window and your budget are constraints the plan respects. They are not the reason the plan exists.

Sixty percent of a window unused every day is capacity that could run the refactor, the test backfill, the migration dry-run in the background. The value is not "spend more"; it is have more done. A suggestion about work, never a nudge about tokens.

my work · tuesday6 rows · 2 need you

when	ask	size	note
now	SSO redirect loop on the Acme tenant	M	worth high · $4–11 · Fable
now	Review PR #409 from last night	S	signed · 340/340 tests
background	Migrate lint rules to the new config	S	worth low · $1–3 · Haiku
overnight	Backfill tests for billing webhooks	M	worth medium · $9–24 · ½ price
teammate	Rework the export API	L	Dana owns billing · headroom
held	Rename the internal enum	S	worth unstated · $3–8

the wallan exit, not a dead terminal

At 90% the seat tells you. At 100% you have six exits, priced.

You know what it takes before you pick a model.

Time, tokens and where it lands in your window, per model. The status line shows the reset, not just a percentage. At 90% the seat tells you inside your own session.

status line

sled  ▮▮▮▮▮▮▮▮▮░  90% · resets 1h 12m · credits $14
this  ENG-482 · M · $4–11 · 20–55 min · cap $20
fits  Fable (band) · Sonnet +40% time ⅕ cost
next  ENG-479 fits overnight · Dana has headroom for ENG-471

The exit is the workspace's policy, not a person at 2 am.

Continue on the platform key within your allowance, hold until the reset, run it overnight at half price, switch model, hand off to a buddy, or defer so it runs itself at the reset. The snapshot travels with the work; nothing you were in the middle of is lost.

the wall · 5-hour bar at 100%

Wall: usage limit reached. Session snapshotted.
Provider: Weekly limit reached. Resets Thursday 9:00 am.
Remaining work: $4.10–9.30 · 20–55 min · 82%
1  Continue now       platform key · allowance left $38.00
2  Hold until reset   free · resumes 9:00 am EDT · saves ≈ $6.70 vs continuing now
3  Run overnight      batch tier · ½ price
4  Switch model       sonnet 5 · platform key · $0.90–2.10 · same time · rest of session
5  Hand off           to Dana (idle 42m) · Slack + /sled pickup
6  Defer to reset     runs itself 9:00 am EDT · same model · full price · $4.10–9.30
Reply /sled 1, /sled 2, /sled 3, /sled 4, /sled 5 or /sled 6.   wall jd7k2m9qa41p
> /sled 1
Continuing on the platform key. Worth high · policy allows · $4.10–9.30 against $38.00. Logged.

Air cover. The log says the policy chose.

"You spent too much" becomes "the policy chose to continue", and the audit log agrees. Nobody justifies a token bill after the fact.

audit log · workspace

14:02  ENG-482 · wall · seat jake@ at 100%
14:02  policy continue-under-threshold · worst case $11 < $20 · allowance ok
14:02  continued on the platform key · chosen by policy, not by a person
14:36  landed · $7 actual · receipt posted

what you getnine things

Agent work you can read

What the agent did, on which model, at what cost, against the estimate you started with, signed. How you defend the work later.

A wall is an exit

Hand off, hold, switch, overnight, continue within an allowance. The snapshot travels with the work.

Before, not after

Time, tokens and where it lands in your window, per model, before you pick one.

Air cover

At the wall the exit is policy under the org's rules, and the audit log says so.

Proof

The receipt signs the model, the time, the outputs. In a compliance review that protects you as much as the org.

A buddy when your team is asleep

Hand-offs you can accept, decline or bounce with the reason attached; quiet hours; a DM instead of a channel.

Coaching only you see

Which model you finish in band on, where sessions churn context, when you tend to hit the wall. Yours. Not a manager's.

Have more done

"Here is a task on your list that fits an overnight run; you haven't tried one." The unused window, put to work.

You can leave

sleds uninstall removes every trace. Your tracker keeps the estimates and actuals.

what sleds never doesa product rule, not a tagline

If a VP screenshotted a screen and you would feel graded, it does not ship in that shape.

never	instead
Ranks people	The spread: a distribution, no names. No leaderboard, no sorted per-person spend list, no export that becomes one.
Reads low spend as a signal about you	It is a fact about routing and work. The words are headroom, routed, shifted. No word in the product describes a person as under-used.
Shows your numbers to a manager by default	Leads see aggregates by team and area. Per-person spend is a grant an owner turns on, recorded in the audit log, visible to you when it is on.
Ranks fit or pairing	Fit goes to you and to the dispatcher. It picks a seat and says why; it does not publish a table.
Imposes work	Every routed hand-off is offered with its reason. Decline without explaining; the sender is told either way.
Grades you with coaching	Coaching is yours. A line that can be screenshotted as an assessment of a person does not ship.
Becomes an HR tool	No performance-review use, no productivity score, no time tracking. If a customer asks, the answer is no.

Why this is also the business: the individual installs sleds because it plans their week and gets them through walls. The team follows because hand-offs and the spread need everyone on it. The enterprise buys because the record, the policy and the receipt are already there — and because the engineers already trust it.

installone line · accept the handles it finds

Nothing changes in your session until you ask for an estimate or hit a limit.

sleds init captures your own handles — git email, GitHub login, the agent's account — and Enter accepts them. It backs up your agent settings once, points the agent at your seat, installs the status line, the /sled skill and the hooks. sleds doctor says what is wired, including your cockpit. sleds off returns the agent to your seat directly.

The CLI, hooks, status line and proxy client are open source. "Your laptop, your install" is easier to believe when you can read it.

terminal

$ curl -fsSL sleds.ai/install | sh
✓ sleds 0.9 · mac arm64
$ sleds init
  git email      dana@acme.dev        ↵ accept
  github         dana-acme            ↵ accept
  agent          claude code · dana@  ↵ accept
  cockpit        cmux                 detected
✓ status line · /sled skill · hooks · settings backed up once
$ sleds doctor
  seat ok · tracker linear · cockpit cmux · key workspace

Free for one engineer.
Forever.

$curl -fsSL sleds.ai/install | sh▍

claude code · codex cli · cursor · cmux · orca