Run it beside the old tool

# Not live yet (gaps #2)
curl -X POST "https://api.sb/leaseline.app/campaigns?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"run-beside"}'

 1import { $ } from 'startup-builder'
 2
 3return await $.Campaign.create({ play: 'run-beside' }, { mode: 'try' })   // dry run on the day-0 Startup, 2026-08-24 · Not live yet, gaps #2

npx startup-builder do Campaign.create --json '{"play":"run-beside"}' --try   # Not live yet (gaps #2)

{
  "name": "do",
  "arguments": {
    "code": "return await $.Campaign.create({ play: 'run-beside' }, { mode: 'try' })   // dry run on the day-0 Startup, 2026-08-24 · Not live yet, gaps #2",
    "mode": "try"
  }
}

 1{
 2  "type": "https://startupbuilder.dev/errors/refused",
 3  "title": "Refused",
 4  "status": 422,
 5  "detail": "run-beside can’t start yet: 1 check fails and 2 warn.",
 6  "checks": [
 7    {
 8      "check": "use-when",
 9      "status": "pass",
10      "observed": "41 Customers; 6 switchers a month name an old tool at intake",
11      "doc": "/plays/run-beside#check-use-when",
12      "fix": "Keep 4 or more switchers a month, or use standard onboarding.",
13      "then": "Metric.query"
14    },
15  …

curl "https://api.sb/leaseline.app/events?where=name%20in%20(%22Order.paid%22%2C%20%22Integration.connected%22%2C%20%22Integration.synced%22%2C%20%22Monitor.checked%22%2C%20%22Integration.revoked%22%2C%20%22Subscription.unsubscribed%22)&limit=1" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

# Not live yet (gaps #7)
curl -X POST "https://api.sb/leaseline.app/triggers/fire?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"like":"event_9910"}'

# Not live yet (gaps #7)
curl "https://api.sb/leaseline.app/decisions?where=on%20%3D%20%22customer_71%22" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

# Not live yet (gaps #7)
curl "https://api.sb/leaseline.app/actions?where=on%20%3D%20%22customer_71%22" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY"

 1import { $ } from 'startup-builder'
 2
 3const { events: [last] } = await $.Event.list({ where: 'name in ("Order.paid", "Integration.connected", "Integration.synced", "Monitor.checked", "Integration.revoked", "Subscription.unsubscribed")', limit: 1 })   // { play } compiled to this Play’s Trigger Events
 4const { subject } = await $.Trigger.fire({ like: last.$id }, { mode: 'try' })   // the same Event again, in try mode · Not live yet, gaps #7
 5const { decisions: [decision] } = await $.Decision.list({ where: `on = "${subject}"` })   // what it chose, and why · Not live yet, gaps #7
 6const { actions } = await $.Action.list({ where: `on = "${subject}"` })   // what it did, with its checks · Not live yet, gaps #7
 7return { decision, actions }

npx startup-builder do Event.list --json '{"where":"name in (\"Order.paid\", \"Integration.connected\", \"Integration.synced\", \"Monitor.checked\", \"Integration.revoked\", \"Subscription.unsubscribed\")","limit":1}'
npx startup-builder do Trigger.fire --json '{"like":"event_9910"}' --try   # Not live yet (gaps #7)
npx startup-builder do Decision.list --json '{"where":"on = \"customer_71\""}'   # Not live yet (gaps #7)
npx startup-builder do Action.list --json '{"where":"on = \"customer_71\""}'   # Not live yet (gaps #7)

{
  "name": "do",
  "arguments": {
    "code": "const { events: [last] } = await $.Event.list({ where: 'name in (\"Order.paid\", \"Integration.connected\", \"Integration.synced\", \"Monitor.checked\", \"Integration.revoked\", \"Subscription.unsubscribed\")', limit: 1 })   // { play } compiled to this Play’s Trigger Events\nconst { subject } = await $.Trigger.fire({ like: last.$id }, { mode: 'try' })   // the same Event again, in try mode · Not live yet, gaps #7\nconst { decisions: [decision] } = await $.Decision.list({ where: `on = \"${subject}\"` })   // what it chose, and why · Not live yet, gaps #7\nconst { actions } = await $.Action.list({ where: `on = \"${subject}\"` })   // what it did, with its checks · Not live yet, gaps #7\nreturn { decision, actions }",
    "mode": "try"
  }
}

 1{
 2  "decision": {
 3    "$id": "https://api.sb/leaseline.app/decisions/decision_866",
 4    "$type": "Decision",
 5    "choice": "book-the-cutover",
 6    "reason": "Ashgrove Lettings’ run read 0 open differences for 5 run days over 170 compared notices; all 4 ours patterns are fixed",
 7    "stakes": "medium",
 8    "options": [
 9      "start-the-run",
10      "standard-onboarding",
11      "fix-ours",
12      "ask-the-choice",
13      "meet-on-a-stall",
14      "book-the-cutover",
15  …

action_2 is the cutover ask to Ashgrove’s admin, waiting for your approval because the stakes are medium. It asks them for a date; nothing Leaseline makes reaches their live records before it.

≈ +$230 ARR a month at 6, kept

Worth

40 min week 1, then 10 a week

Your time

$1.10 per run, ~$7 a month at 6

Spend

Day 19: the first cutover booked

First signal

4+ switchers a month, 20+ items a day

Use it when

Under 4: onboard-customer · no read-only access: switch-kit

Not yet when

Worth’s formula: about +22 cutovers per 100 switchers (Other Startups’ results; no lift of your own yet) × 6 switchers a month × 0.5 kept to month 3 that would have left × $347 Pro ARR ≈ +$229 of churn avoided. Under First signal: At 6 switchers a month, 90 days can tell only +40 per 100 or more, so your file starts at holdout: 0 and reads before-after.

Start it this weekCopy Anchor Link

npx startup-builder add play run-beside   # 1 · writes plays/run-beside.mdx (Not live yet, gaps #5)

# 2 · dry run; below, its 422 on 2026-08-24
curl -X POST "https://api.sb/leaseline.app/campaigns?mode=try" \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"run-beside"}'

# 3 · keep it, at You approve the first wave
curl -X POST https://api.sb/leaseline.app/campaigns \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"play":"run-beside"}'

# on
curl -X PATCH https://api.sb/leaseline.app/campaigns/run-beside \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"endsAt":null}'

# off: open runs end with their report; switchers get standard onboarding
curl -X PATCH https://api.sb/leaseline.app/campaigns/run-beside \
  -H "Authorization: Bearer $STARTUP_BUILDER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"endsAt":"now"}'

{
  "type": "https://startupbuilder.dev/errors/refused",
  "title": "Refused",
  "status": 422,
  "detail": "run-beside can’t start yet: 1 check fails and 2 warn.",
  "checks": [
    {
      "check": "use-when",
      "status": "pass",
      "observed": "41 Customers; 6 switchers a month name an old tool at intake",
      "doc": "/plays/run-beside#check-use-when",
      "fix": "Keep 4 or more switchers a month, or use standard onboarding.",
      "then": "Metric.query"
    },
    {
      "check": "read-only",
      "status": "fail",
      "observed": "the old-tool connection asks for write access",
      "doc": "/plays/run-beside#check-read-only",
      "fix": "Connect the old tool with its read-only export scope instead.",
      "then": "Integration.connect"
    },
    {
      "check": "baseline",
      "status": "warn",
      "observed": "3 of 8 switchers retired the old tool within 45 days, in 90 days",
      "doc": "/plays/run-beside#check-baseline",
      "fix": "n 8: it reads before-after, not proof, until volume allows a held-back half.",
      "then": "Metric.query"
    },
    {
      "check": "role-unstaffed",
      "status": "warn",
      "roles": [
        "success-manager",
        "customer-lead"
      ],
      "owned": 4,
      "doc": "/plays/run-beside#check-role-unstaffed",
      "fix": "Staff the CSM, or leave it: stalls come to you, and Priya’s 4 get drafts.",
      "then": "Role.staff"
    }
  ]
}

Each run costs about $1.10 in shadow compute and model calls on differences, and each update $0.04 (Worker cost). The miss rule refunds the first month: 1 so far, $29. That comes to about $7 a month at 6 switchers, plus refunds.

When it startsCopy Anchor Link

{
  "$id": "https://api.sb/leaseline.app/events/event_9910",
  "$type": "Event",
  "name": "Monitor.checked",
  "at": "2026-10-18T07:00:00Z",
  "seq": 9910,
  "data": {
    "customer": "customer_71",
    "run": "run_9",
    "open": 0,
    "streak": 5,
    "compared": 170,
    "oursFixed": 4,
    "day": 23
  }
}

npx startup-builder trigger Monitor.checked --like event_9910   # replays the fifth zero day
npx startup-builder clock advance --by 2d   # runs the cutover ask’s day-2 step
npx startup-builder listen Monitor.checked --forward-to http://localhost:3000/hook

Who runs itCopy Anchor Link

CSM · your agent (role:success-manager, worker_5) holds the Key Result. Stalls go to the Head of customers Role, which has no person, so you (Dana Okafor, person, worker_1), and so do day-45 refunds and every cutover ask, which you approve. The 4 Customers AE · Priya Shah (person) owns get drafts in her queue. Your number links to Did it work.

What it reads · What it may call. It reads the old tool through its read-only export and only the fields the run sheet lists; a report view or a login never appears in a Message. Nothing it makes reaches live records before the cutover date the Customer confirms (shadow-hold). Its tools are POST /{namespace}/integrations/connect, POST /{namespace}/issues/{issue}/triage, POST /{namespace}/changes/{change}/review, POST /{namespace}/conversations/{conversation}/answer, POST /{namespace}/meetings/book, POST /{namespace}/tasks/{task}/assign, POST /{namespace}/integrations/{integration}/sync, POST /{namespace}/invoices/{invoice}/refund, and any other call becomes an Approval under the Role’s Mandate.

You hold the CSM Key Result: 80 cutovers per 100 switchers by 2026-12-31.        ← worker, metric
On each Trigger, choose one option and say why; a held Customer gets Not now.    ← triggers, chooses, keep-first
Compare due date, amount and recipient on every run day’s notices.               ← your file, line 4
List the differences you caused first, each with its evidence.                   ← copyRules 2, 3
Say zero open differences over a count of items and days, never zero errors.     ← copyRules 4
A stall gets one Meeting with Dana; day 45 with no streak refunds the month.     ← your file, line 5

The slow loop. The weekly Decisions kept the run sheet’s fields and the 5-day streak, and no Change was merged in the window, so there is no replay or read-back to print. Each ours fix shipped through POST /{namespace}/changes/{change}/review against the Product’s eval set, never as a change to this Play.

Bring your own agent (folded) starts at It does it, you can stop it for difference classing when it matches the seated agent on the eval suite; cutover asks stay at You approve each one. POST /{namespace}/roles/{role}/staff seats it.

npx startup-builder eval --scenario run-beside-11 --agent "claude -p" --seat success-manager   # seated: 7 of 9 cut over, every ours fix eval-gated · yours: 6 of 9, 1 difference classed theirs that was ours

What it can do nextCopy Anchor Link

When it hands to youCopy Anchor Link

Stalls, cutover asks and day-45 refunds come to you: the CSM and Head of customers Roles have no person. One Brief per Role, as captured (brief is Not live yet, gaps #7).

[
  {
    "$id": "https://api.sb/leaseline.app/tasks/task_871",
    "$type": "Task",
    "assignee": "worker_1",
    "due": "2026-10-09T17:00:00Z",
    "on": "customer_74",
    "brief": {
      "who": "Rowan Place, Pro since Sep 22",
      "said": "open differences flat at 7 for 5 run days",
      "chose": "meet-on-a-stall",
      "why": "a stall gets one Meeting",
      "open": "7 choice differences on pro-rated rent rounding",
      "never": [
        "zero errors",
        "a claim about the old tool beyond this run"
      ]
    }
  },
  {
    "$id": "https://api.sb/leaseline.app/approvals/approval_140",
    "$type": "Approval",
    "verdict": null,
    "deadline": "2026-10-20T17:00:00Z",
    "draft": "Your run has read zero open differences for 5 run days over 170 notices. Which date should we switch on?",
    "route": "book-the-cutover",
    "checks": [
      {
        "check": "shadow-hold",
        "status": "pass"
      },
      {
        "check": "claims-resolve",
        "status": "pass"
      }
    ]
  }
]

You answer with POST /{namespace}/tasks/{task}/claim (Take this one), POST /{namespace}/tasks/{task}/return with a reason, POST /{namespace}/actions/{action}/redirect, POST /{namespace}/approvals/{approval}/approve or POST /{namespace}/approvals/{approval}/refuse, each a one-call window. A cancel ask ends the run within 1 hour and never waits on you (exit-never-waits). A Task nobody claims by its due time goes up a Role through POST /{namespace}/tasks/{task}/escalate, and person-promise fires.

On time: you, 10 of 10 within their due time, median 3 h, n 10, ■□□□ Too early to tell beside Other Startups’ 94 % on time. Your queue: in week 1, 4 approvals and 1 weekly review; then about 1 cutover ask and 1 stall every two weeks.

The fixed line: POST /{namespace}/deals/{deal}/assign sets owner to the receiving person, the agent is never a Deal’s owner, and Meetings land on the owner’s calendar under the owner’s name.

When it won’t sendCopy Anchor Link

The captured refusal, then the Stop that fired as a Campaign.paused reason (Not live yet, gaps #2); the update that sent is the Linden Court thread in Example.

{
  "type": "https://startupbuilder.dev/errors/refused",
  "title": "Refused",
  "status": 422,
  "detail": "The update to Linden Court did not send: it says their old tool makes errors.",
  "checks": [
    {
      "check": "claims-resolve",
      "status": "fail",
      "doc": "/plays/run-beside#check-claims-resolve",
      "fix": "Say zero open differences over the items and days counted.",
      "then": "Content.draft"
    },
    {
      "check": "keep-first",
      "status": "pass"
    }
  ]
}

[
  {
    "reason": {
      "limit": "stall",
      "observed": "7 open, flat",
      "bound": "5 run days",
      "window": "run_11, Rowan Place"
    },
    "doc": "/plays/run-beside#stop-stall",
    "task": "task_871"
  }
]

Did it workCopy Anchor Link

At 6 switchers a month no held-back half reads within 90 days, so the basis is before-after (rule o): the columns read Now and Before, and nothing prints as Added. Each figure’s call is POST /{namespace}/metrics/query by node; the leaf and every node above it are Not live yet (gaps #4).

{
  "$id": "decisions/decision_866",
  "$type": "Decision",
  "choice": "book-the-cutover",
  "reason": "Ashgrove Lettings’ run read 0 open differences for 5 run days over 170 compared notices; all 4 ours patterns are fixed",
  "stakes": "medium",
  "options": [
    "start-the-run",
    "standard-onboarding",
    "fix-ours",
    "ask-the-choice",
    "meet-on-a-stall",
    "book-the-cutover",
    "end-and-refund",
    "not-now"
  ],
  "by": "worker_5",
  "on": "customer_71",
  "expected": "the old tool off by the cutover date Ashgrove gives, and still paying in month 3",
  "reopensWhen": "no Integration.revoked for customer_71 within 14d of the cutover date"
}

Before against after reads one window against the 90 days before it, scaled to 56 days; it is not proof, and Other Startups’ results print beside it. The leaf counts switching Customers who start a run and retire the old tool within 45 days of their first paid Order.

How it moves it: each switcher saw every difference between the two tools listed and fixed, ours first, and reached a zero streak in a median of 21 days.

Guardrails: before against after, the family pair under churned-arr: contraction-arr, $0 from switches down against $347 before, and contribution-margin, run cost $1.10 a Customer against $29 a month.

{
  "$id": "https://api.sb/leaseline.app/campaigns/run-beside",
  "$type": "Campaign",
  "play": "play:run-beside@1",
  "path": [
    "cutover-rate",
    "month-3-retention",
    "churned-arr",
    "net-new-arr",
    "arr"
  ],
  "experiment": {
    "metric": "cutover-rate",
    "unit": "per-100",
    "heldBack": 0,
    "decision": null
  },
  "readout": {
    "value": null,
    "n": 9,
    "maturing": 2,
    "interval": null,
    "readability": "inadmissible",
    "basis": "before-after",
    "pooled": true
  },
  "ledger": {
    "spend": 29,
    "workerCost": 13,
    "hours": 1.6,
    "pipelineArr": 0,
    "wonArr": 0,
    "projectedArr": 229,
    "paybackDays": null
  }
}

Outside benchmark: this Play’s catalog record carries no sourced range yet, so Other Startups’ results stand in: +22 cutovers per 100 switchers, ■■■□ Likely, n 1,140.

Before it · After it. Before it: switch-kit and paid-pilot bring the switchers. After it: review-requests asks at the cutover, customer-referrals hears the Customers who switched, and usage-upsell takes them past their Limit; keep-first holds a Customer at risk.

ExampleCopy Anchor Link

Leaseline ran it from 2026-08-24 to 2026-10-19 (56 days): 14 switchers paid, 3 went to standard onboarding, and 11 started a run; 9 have passed day 45. Read every captured thread at ?threads=all; three follow the tables.

Leaseline → Linden Court’s admin, Sep 13 08:00 · Day 19 of your run: 0 open differences for 5 run days, over 168 renewal notices compared on due date, amount and recipient.¹ The 4 we caused are fixed and listed in your report.² Which date should we switch on? *Signed with the Role’s disclosure line.*³

Linden Court, Sep 13 10:20 · “Let’s switch on Oct 1.” The old tool went off on Oct 1; Linden Court is still paying.

- Good news: we found 6 errors in your old tool that Leaseline gets right.
+ Day 10: 6 open differences, all on the old tool’s side, each listed with the notice it came from.

Rowan Place, 2026-10-08 07:00 · 7 open differences flat for 5 run days, all on pro-rated rent rounding · The agent, 07:01 · paused the weekly note and assigned the stall to you (task_871), due 17:00 · Dana, 09:30 · claimed it and booked a 20-minute call for Oct 9; the rounding became a Setting.

¹ Zero open differences over a stated count of items and days, never zero errors. ² The differences you caused, listed first, each with its evidence. ³ The fixed signature line, compiled from the Role’s disclosure Setting.

Your settingsCopy Anchor Link

---
holdout: 0        # 6 switchers a month is too few to hold half back; it reads before-after
---
Run beside any switcher with read-only access and 20 or more notices a day.
Compare due date, amount and recipient; book the switch after 5 zero days over 100 notices.
A stall gets one call with Dana; day 45 with no streak refunds the first month.

$ npx startup-builder show plays/run-beside
holdout 0 · 20 notices a day · 3 fields · 5 zero days over 100 · stall call · day-45 refund     your file  plays/run-beside.mdx:2,4,5,6 (Ask 2026-08-24)
worker role:success-manager, owns never · chooses 8 (6 rule, 2 read) · sequences parallel-run-updates, cutover-asks · approval first-send, every-send · checks owned-by-a-person, keep-first, customer-ask-cap, exit-never-waits, read-only, shadow-hold, field-scope, regression, copy-deletion, claims-resolve · stops 3 · placements gmail, outlook + the difference report    catalog  play:run-beside@1 (Not live yet, gaps #5)

Every key from the catalogCopy Anchor Link

Compiled, not written: edit the Play record run-beside and its Leaseline run, then regenerate. Every step runs from a sentence or a line of code. In your words: the CRM words you know, and where each lives here.